[Model] Optimize MiniMax-H3 DLO component lifecycle

Signed-off-by: hsliuustc0106 <liuhongsheng4@huawei.com>

.buildkite/amd/test-amd-merge.yml

@@ -17,7 +17,7 @@ steps:
         - export VLLM_ROCM_USE_AITER=0
         # ignore test_teacache_extractors.py because it use rocm gemm kernel from vLLM
         # that is not supported on CPU
-        - "pytest -sv tests/model_executor -m 'core_model and cpu' --num-shards=$$BUILDKITE_PARALLEL_JOB_COUNT --shard-id=$$BUILDKITE_PARALLEL_JOB --cov=vllm_omni --cov-report=term-missing:skip-covered --cov-report=xml:coverage-model-executor-$$BUILDKITE_PARALLEL_JOB.xml"
+        - "pytest -sv tests/model_executor -m 'core_model and cpu' --num-shards=$$BUILDKITE_PARALLEL_JOB_COUNT --shard-id=$$BUILDKITE_PARALLEL_JOB"
 
     - label: "Simple · Diffusion Test · Shard %N/%t"
       agent_pool: mi300_1
@@ -29,7 +29,7 @@ steps:
         - export VLLM_ROCM_USE_AITER=0
         # ignore test_teacache_extractors.py because it use rocm gemm kernel from vLLM
         # that is not supported on CPU
-        - "pytest -sv tests/diffusion -m 'core_model and cpu' --ignore=tests/diffusion/cache/test_teacache_extractors.py --num-shards=$$BUILDKITE_PARALLEL_JOB_COUNT --shard-id=$$BUILDKITE_PARALLEL_JOB --cov=vllm_omni --cov-report=term-missing:skip-covered --cov-report=xml:coverage-diffusion-$$BUILDKITE_PARALLEL_JOB.xml"
+        - "pytest -sv tests/diffusion -m 'core_model and cpu' --ignore=tests/diffusion/cache/test_teacache_extractors.py --num-shards=$$BUILDKITE_PARALLEL_JOB_COUNT --shard-id=$$BUILDKITE_PARALLEL_JOB"
 
     - label: "Simple · Engine&Entrypoints Test"
       agent_pool: mi300_1
@@ -38,7 +38,7 @@ steps:
       grade: Blocking
       commands:
         - export VLLM_ROCM_USE_AITER=0
-        - "pytest -sv tests/entrypoints tests/engine -m 'core_model and cpu' --cov=vllm_omni --cov-report=term-missing:skip-covered --cov-report=xml"
+        - "pytest -sv tests/entrypoints tests/engine -m 'core_model and cpu'"
 
     - label: "Simple · Other Test"
       agent_pool: mi300_1
@@ -49,7 +49,7 @@ steps:
         - export VLLM_ROCM_USE_AITER=0
         # ignore test_teacache_extractors.py because it use rocm gemm kernel from vLLM
         # that is not supported on CPU
-        - "pytest -sv tests/ -m 'core_model and cpu' --ignore=tests/diffusion --ignore=tests/model_executor --ignore=tests/entrypoints --ignore=tests/engine  --ignore=tests/diffusion/cache/test_teacache_extractors.py --cov=vllm_omni --cov-report=term-missing:skip-covered --cov-report=xml"
+        - "pytest -sv tests/ -m 'core_model and cpu' --ignore=tests/diffusion --ignore=tests/model_executor --ignore=tests/entrypoints --ignore=tests/engine  --ignore=tests/diffusion/cache/test_teacache_extractors.py"
 
 - group: ":card_index_dividers: Diffusion Test"
   steps:

.buildkite/amd/test-amd-ready.yml

@@ -17,7 +17,7 @@ steps:
         - export VLLM_ROCM_USE_AITER=0
         # ignore test_teacache_extractors.py because it use rocm gemm kernel from vLLM
         # that is not supported on CPU
-        - "pytest -sv tests/model_executor -m 'core_model and cpu' --num-shards=$$BUILDKITE_PARALLEL_JOB_COUNT --shard-id=$$BUILDKITE_PARALLEL_JOB --cov=vllm_omni --cov-report=term-missing:skip-covered --cov-report=xml:coverage-model-executor-$$BUILDKITE_PARALLEL_JOB.xml"
+        - "pytest -sv tests/model_executor -m 'core_model and cpu' --num-shards=$$BUILDKITE_PARALLEL_JOB_COUNT --shard-id=$$BUILDKITE_PARALLEL_JOB"
 
     - label: "Simple · Diffusion Test"
       agent_pool: mi300_1
@@ -28,7 +28,7 @@ steps:
         - export VLLM_ROCM_USE_AITER=0
         # ignore test_teacache_extractors.py because it use rocm gemm kernel from vLLM
         # that is not supported on CPU
-        - "pytest -sv tests/diffusion -m 'core_model and cpu'  --ignore=tests/diffusion/cache/test_teacache_extractors.py --cov=vllm_omni --cov-report=term-missing:skip-covered --cov-report=xml"
+        - "pytest -sv tests/diffusion -m 'core_model and cpu'  --ignore=tests/diffusion/cache/test_teacache_extractors.py"
 
     - label: "Simple · Engine&Entrypoints Test"
       agent_pool: mi300_1
@@ -37,7 +37,7 @@ steps:
       grade: Blocking
       commands:
         - export VLLM_ROCM_USE_AITER=0
-        - "pytest -sv tests/entrypoints tests/engine -m 'core_model and cpu' --cov=vllm_omni --cov-report=term-missing:skip-covered --cov-report=xml"
+        - "pytest -sv tests/entrypoints tests/engine -m 'core_model and cpu'"
 
     - label: "Simple · Other Test"
       agent_pool: mi300_1
@@ -48,7 +48,7 @@ steps:
         - export VLLM_ROCM_USE_AITER=0
         # ignore test_teacache_extractors.py because it use rocm gemm kernel from vLLM
         # that is not supported on CPU
-        - "pytest -sv tests/ -m 'core_model and cpu' --ignore=tests/diffusion --ignore=tests/model_executor --ignore=tests/entrypoints --ignore=tests/engine  --ignore=tests/diffusion/cache/test_teacache_extractors.py --cov=vllm_omni --cov-report=term-missing:skip-covered --cov-report=xml"
+        - "pytest -sv tests/ -m 'core_model and cpu' --ignore=tests/diffusion --ignore=tests/model_executor --ignore=tests/entrypoints --ignore=tests/engine  --ignore=tests/diffusion/cache/test_teacache_extractors.py"
 
 - label: "Custom Pipeline Test"
   agent_pool: mi300_1

.buildkite/common/scripts/run_cov_split.sh

@@ -16,6 +16,11 @@
 # At least one of --offline/--online is required, so a job with only one mode runs
 # just that half. Total runtime is bounded by the step's timeout_in_minutes.
 #
+# Coverage + artifact upload run only on scheduled weekly E2E sweeps
+# (WEEKLY=1 and BUILDKITE_BRANCH=main — same gate as bootstrap --e2e).
+# All other invocations (PR merge-test, regular main merge, local) run the
+# same combined pytest as the job used before coverage was added.
+#
 # Exits non-zero if any half or the upload failed.
 set -uo pipefail
 
@@ -44,6 +49,17 @@ if ((${#OFFLINE[@]} == 0 && ${#ONLINE[@]} == 0)); then
     exit 2
 fi
 
+_should_run_coverage() {
+    [[ "${WEEKLY:-}" == "1" && "${BUILDKITE_BRANCH:-}" == "main" ]]
+}
+
+if ! _should_run_coverage; then
+    PATHS=("${OFFLINE[@]}" "${ONLINE[@]}")
+    echo "--- pytest (no coverage): ${MODEL_ID}"
+    pytest -s -v "${PATHS[@]}" "${PYTEST_ARGS[@]}"
+    exit $?
+fi
+
 # Two steps can test the same model, and artifacts are scoped to the build (a
 # nightly one carries the ready, merge and nightly tiers at once), so unsuffixed
 # reports would share a path -- a glob download then races to rename onto one file.

.buildkite/common/scripts/upload_pipeline.py

@@ -1,5 +1,6 @@
 #!/usr/bin/env python3
 # SPDX-License-Identifier: Apache-2.0
+# SPDX-FileCopyrightText: Copyright contributors to the vLLM-Omni project
 """Render and optionally upload Buildkite pipeline YAML with diff-aware logic.
 
 Bootstrap mode (``bootstrap-upload-steps.yml``):
@@ -58,11 +59,12 @@ BOOTSTRAP_UPLOAD_IF_KEYS = {
 E2E_GROUP_MARKER = "E2E Test"
 CI_MIRROR_HARDWARES_PATH = ROOT / ".buildkite/common/ci_mirror_hardwares.yml"
 
-CUDA_NIGHTLY_ONLY = (
-    '(build.pull_request.labels includes "nightly-test") || (build.branch == "main" && build.env("NIGHTLY") == "1")'
-)
-NPU_NIGHTLY_ONLY = (
-    '(build.branch == "main" && build.env("NIGHTLY") == "1") || '
+# Bootstrap Buildkite ``if`` expressions.
+# ``*_MAIN_IF``: main + env schedule. ``*_LABEL_IF``: PR label (and/or composed with MAIN).
+# ``*_UPLOAD_IF``: full gate for uploading that child pipeline.
+NIGHTLY_MAIN_IF = 'build.branch == "main" && build.env("NIGHTLY") == "1"'
+NIGHTLY_LABEL_IF = (
+    f"({NIGHTLY_MAIN_IF}) || "
     '(build.branch != "main" && ('
     'build.pull_request.labels includes "nightly-test" || '
     'build.pull_request.labels includes "omni-test" || '
@@ -71,6 +73,19 @@ NPU_NIGHTLY_ONLY = (
     'build.pull_request.labels includes "diffusion-x2v-test"'
     "))"
 )
+WEEKLY_E2E_IF = 'build.branch == "main" && build.env("WEEKLY") == "1"'
+WEEKLY_MAIN_IF = 'build.branch == "main" && (build.env("WEEKLY") == "1" || build.env("NON_CRITICAL") == "1")'
+WEEKLY_LABEL_IF = f'({WEEKLY_MAIN_IF}) || (build.branch != "main" && build.pull_request.labels includes "weekly-test")'
+READY_LABEL_IF = 'build.branch != "main" && build.pull_request.labels includes "ready"'
+MERGE_LABEL_IF = 'build.branch != "main" && build.pull_request.labels includes "merge-test"'
+MERGE_MAIN_IF = (
+    'build.branch == "main" && build.env("NIGHTLY") != "1" && '
+    'build.env("WEEKLY") != "1" && build.env("NON_CRITICAL") != "1"'
+)
+READY_UPLOAD_IF = f"({WEEKLY_E2E_IF}) || ({READY_LABEL_IF})"
+MERGE_UPLOAD_IF = f"({WEEKLY_E2E_IF}) || (({MERGE_MAIN_IF}) || ({MERGE_LABEL_IF}))"
+BOOTSTRAP_DISABLED_IF = "false"
+BOOTSTRAP_ENABLED_IF = "true"
 
 
 # --- Logging ---
@@ -102,74 +117,47 @@ def _format_bootstrap_if(expr: str) -> str:
 
 
 def _compute_bootstrap_if_exprs(*, decision, platform: str) -> dict[str, str]:
-    disabled = "false"
-    nightly_main = 'build.branch == "main" && build.env("NIGHTLY") == "1"'
-    nightly_only = NPU_NIGHTLY_ONLY if platform == "npu" else CUDA_NIGHTLY_ONLY
-
     if platform == "npu":
-        ready_pr = (
-            'build.branch != "main" && ('
-            'build.pull_request.labels includes "npu-test" || '
-            'build.pull_request.labels includes "ready"'
-            ")"
-        )
-        nightly_label_if = nightly_only
-        weekly_label_if = disabled
-        merge_base = disabled
+        ready_upload = READY_LABEL_IF
+        merge_upload = BOOTSTRAP_DISABLED_IF
+        weekly_label_if = BOOTSTRAP_DISABLED_IF
     else:
-        ready_pr = 'build.branch != "main" && build.pull_request.labels includes "ready"'
-        merge_main = 'build.branch == "main" && build.env("NIGHTLY") != "1" && build.env("WEEKLY") != "1"'
-        merge_pr = 'build.branch != "main" && build.pull_request.labels includes "merge-test"'
-        nightly_label_if = (
-            '(build.branch == "main" && build.env("NIGHTLY") == "1") || '
-            '(build.branch != "main" && ('
-            'build.pull_request.labels includes "nightly-test" || '
-            'build.pull_request.labels includes "omni-test" || '
-            'build.pull_request.labels includes "tts-test" || '
-            'build.pull_request.labels includes "diffusion-x2iat-test" || '
-            'build.pull_request.labels includes "diffusion-x2v-test"'
-            "))"
-        )
-        weekly_label_if = (
-            '(build.branch == "main" && build.env("WEEKLY") == "1") || '
-            '(build.branch != "main" && build.pull_request.labels includes "weekly-test")'
-        )
-        merge_base = f"({nightly_main}) || (({merge_main}) || ({merge_pr}))"
-
-    ready_base = f"({nightly_main}) || ({ready_pr})"
-    weekly_main = 'build.branch == "main" && build.env("WEEKLY") == "1"'
+        ready_upload = READY_UPLOAD_IF
+        merge_upload = MERGE_UPLOAD_IF
+        weekly_label_if = WEEKLY_LABEL_IF
 
     if decision.skip_all:
         # Docs / skip-mark only: no PR-label escape hatch. Main scheduled
-        # NIGHTLY=1 still runs L4 + L2/L3 (--e2e); WEEKLY=1 still runs L5.
-        image_expr = f"({nightly_main}) || ({weekly_main})" if platform == "cuda" else nightly_main
-        ready_expr = nightly_main
-        merge_expr = nightly_main if platform == "cuda" else disabled
-        nightly_expr = nightly_main
-        weekly_expr = weekly_main if platform == "cuda" else disabled
+        # NIGHTLY=1 still runs L4; WEEKLY=1 / NON_CRITICAL=1 still run L5.
+        # main+WEEKLY=1 also uploads L2/L3 (those steps then pass --e2e).
+        image_expr = f"({NIGHTLY_MAIN_IF}) || ({WEEKLY_MAIN_IF})" if platform == "cuda" else NIGHTLY_MAIN_IF
+        ready_expr = WEEKLY_E2E_IF if platform == "cuda" else BOOTSTRAP_DISABLED_IF
+        merge_expr = WEEKLY_E2E_IF if platform == "cuda" else BOOTSTRAP_DISABLED_IF
+        nightly_expr = NIGHTLY_MAIN_IF
+        weekly_expr = WEEKLY_MAIN_IF if platform == "cuda" else BOOTSTRAP_DISABLED_IF
     elif decision.skip_l2_l3:
         l2_enabled = decision.is_run("npu", "l2") if platform == "npu" else decision.is_run("cuda", "l2")
         l3_enabled = platform == "cuda" and decision.is_run("cuda", "l3")
 
-        ready_expr = ready_base if l2_enabled else disabled
-        merge_expr = merge_base if l3_enabled else disabled
-        nightly_expr = nightly_label_if
-        weekly_expr = weekly_label_if if platform == "cuda" else disabled
+        ready_expr = ready_upload if l2_enabled else BOOTSTRAP_DISABLED_IF
+        merge_expr = merge_upload if l3_enabled else BOOTSTRAP_DISABLED_IF
+        nightly_expr = NIGHTLY_LABEL_IF
+        weekly_expr = weekly_label_if if platform == "cuda" else BOOTSTRAP_DISABLED_IF
 
-        image_parts = [f"({nightly_label_if})"]
+        image_parts = [f"({NIGHTLY_LABEL_IF})"]
         if platform == "cuda":
             image_parts.append(f"({weekly_label_if})")
         if l2_enabled:
-            image_parts.insert(0, f"({ready_base})")
+            image_parts.insert(0, f"({ready_upload})")
         if l3_enabled:
-            image_parts.insert(1 if l2_enabled else 0, f"({merge_base})")
+            image_parts.insert(1 if l2_enabled else 0, f"({merge_upload})")
         image_expr = " || ".join(image_parts)
     else:
-        image_expr = "true"
-        ready_expr = ready_base
-        merge_expr = merge_base if platform == "cuda" else disabled
-        nightly_expr = nightly_label_if
-        weekly_expr = weekly_label_if if platform == "cuda" else disabled
+        image_expr = BOOTSTRAP_ENABLED_IF
+        ready_expr = ready_upload
+        merge_expr = merge_upload if platform == "cuda" else BOOTSTRAP_DISABLED_IF
+        nightly_expr = NIGHTLY_LABEL_IF
+        weekly_expr = weekly_label_if if platform == "cuda" else BOOTSTRAP_DISABLED_IF
 
     return {
         "image": _format_bootstrap_if(image_expr),

.buildkite/cuda/bootstrap-upload-steps.yml

@@ -8,28 +8,28 @@ steps:
     agents:
       queue: "cpu_queue_premerge"
 
-  # L2 Test — PR ready label (diff-aware) or main+NIGHTLY=1 (--e2e, full E2E group)
+  # L2 Test — PR ready label (diff-aware) or main+WEEKLY=1 (--e2e, full E2E group)
   - label: "Upload Ready Pipeline"
     depends_on: image-build
     key: upload-ready-pipeline
     commands:
       - |
         EXTRA=""
-        if [ "${NIGHTLY:-}" = "1" ] && [ "${BUILDKITE_BRANCH}" = "main" ]; then
+        if [ "${WEEKLY:-}" = "1" ] && [ "${BUILDKITE_BRANCH}" = "main" ]; then
           EXTRA="--e2e"
         fi
         python3 .buildkite/common/scripts/upload_pipeline.py --upload $$EXTRA .buildkite/cuda/test-ready.yml
     agents:
       queue: "cpu_queue_premerge"
 
-  # L3 Test — PR merge-test label (diff-aware) or main+NIGHTLY=1 (--e2e, full E2E group)
+  # L3 Test — PR merge-test label (diff-aware) or main+WEEKLY=1 (--e2e, full E2E group)
   - label: "Upload Merge Pipeline"
     depends_on: image-build
     key: upload-merge-pipeline
     commands:
       - |
         EXTRA=""
-        if [ "${NIGHTLY:-}" = "1" ] && [ "${BUILDKITE_BRANCH}" = "main" ]; then
+        if [ "${WEEKLY:-}" = "1" ] && [ "${BUILDKITE_BRANCH}" = "main" ]; then
           EXTRA="--e2e"
         fi
         python3 .buildkite/common/scripts/upload_pipeline.py --upload $$EXTRA .buildkite/cuda/test-merge.yml
@@ -45,7 +45,7 @@ steps:
     agents:
       queue: "cpu_queue_premerge"
 
-  # L5 Test — main+WEEKLY=1 (scheduled), or PR with weekly-test label
+  # L5 Test — main+WEEKLY=1 / NON_CRITICAL=1 (scheduled), or PR with weekly-test label
   - label: "Upload Weekly Pipeline"
     depends_on: image-build
     key: upload-weekly-pipeline

.buildkite/cuda/test-merge.yml

@@ -9,23 +9,27 @@ steps:
     depends_on: upload-merge-pipeline
     steps:
       - label: "Simple · Model Executor Test"
+        timeout_in_minutes: 20
         commands:
-          - "timeout 20m pytest -sv tests/model_executor -m 'core_model and cpu' --cov=vllm_omni --cov-report=term-missing:skip-covered --cov-report=xml"
+          - "pytest -sv tests/model_executor -m 'core_model and cpu'"
         mirror_hardwares: l4_1
 
       - label: "Simple · Diffusion Test"
+        timeout_in_minutes: 20
         commands:
-          - "timeout 20m pytest -sv tests/diffusion -m 'core_model and cpu' --cov=vllm_omni --cov-report=term-missing:skip-covered --cov-report=xml"
+          - "pytest -sv tests/diffusion -m 'core_model and cpu'"
         mirror_hardwares: l4_1
 
       - label: "Simple · Engine&Entrypoints Test"
+        timeout_in_minutes: 20
         commands:
-          - "timeout 20m pytest -sv tests/entrypoints tests/engine -m 'core_model and cpu' --cov=vllm_omni --cov-report=term-missing:skip-covered --cov-report=xml"
+          - "pytest -sv tests/entrypoints tests/engine -m 'core_model and cpu'"
         mirror_hardwares: l4_1
 
       - label: "Simple · Other Test"
+        timeout_in_minutes: 20
         commands:
-          - "timeout 20m pytest -sv tests/ -m 'core_model and cpu' --ignore=tests/diffusion --ignore=tests/model_executor --ignore=tests/entrypoints --ignore=tests/engine --cov=vllm_omni --cov-report=term-missing:skip-covered --cov-report=xml"
+          - "pytest -sv tests/ -m 'core_model and cpu' --ignore=tests/diffusion --ignore=tests/model_executor --ignore=tests/entrypoints --ignore=tests/engine"
         mirror_hardwares: l4_1
 
   - group: ":card_index_dividers: Diffusion Test"
@@ -76,6 +80,9 @@ steps:
           - tests/e2e/online_serving/test_qwen3_tts_customvoice.py
           - tests/e2e/offline_inference/test_qwen3_tts_customvoice.py
           - vllm_omni/model_executor/models/qwen3_tts/
+          - vllm_omni/model_executor/models/common/qwen3_code_predictor.py
+          - vllm_omni/model_executor/models/common/snake_activation.py
+          - vllm_omni/model_executor/models/common/alias_free_activation.py
           - vllm_omni/model_executor/stage_input_processors/qwen3_tts.py
           - vllm_omni/deploy/qwen3_tts.yaml
         commands:
@@ -92,6 +99,9 @@ steps:
           - tests/e2e/online_serving/test_qwen3_tts_base.py
           - tests/e2e/offline_inference/test_qwen3_tts_base.py
           - vllm_omni/model_executor/models/qwen3_tts/
+          - vllm_omni/model_executor/models/common/qwen3_code_predictor.py
+          - vllm_omni/model_executor/models/common/snake_activation.py
+          - vllm_omni/model_executor/models/common/alias_free_activation.py
           - vllm_omni/model_executor/stage_input_processors/qwen3_tts.py
           - vllm_omni/deploy/qwen3_tts.yaml
           - .buildkite/common/scripts/run_cov_split.sh
@@ -109,27 +119,17 @@ steps:
               -- -m 'advanced_model and cuda' --run-level 'advanced_model'
         mirror_hardwares: l4_1
 
-      - label: "TTS · Higgs-Audio-V3 Test"
-        source_file_dependencies:
-          - tests/e2e/online_serving/test_higgs_audio_v3.py
-          - vllm_omni/model_executor/models/higgs_audio_v3/
-          - vllm_omni/model_executor/stage_input_processors/higgs_audio_v3.py
-          - vllm_omni/deploy/higgs_multimodal_qwen3.yaml
-        commands:
-          - |
-            timeout 20m bash -c "
-              export VLLM_WORKER_MULTIPROC_METHOD=spawn
-              export VLLM_USE_DEEP_GEMM=0
-              export VLLM_MOE_USE_DEEP_GEMM=0
-              pytest -s -v tests/e2e/online_serving/test_higgs_audio_v3.py -m 'advanced_model' --run-level 'advanced_model'
-            "
-        mirror_hardwares: l4_1
-
       - label: "Omni · Qwen3-Omni Test"
         source_file_dependencies:
           - tests/e2e/offline_inference/test_qwen3_omni.py
           - tests/e2e/online_serving/test_qwen3_omni.py
+          - tests/helpers/runtime.py
+          - vllm_omni/entrypoints/openai/serving_chat.py
           - vllm_omni/model_executor/models/qwen3_omni/
+          - vllm_omni/model_executor/models/common/qwen3_code_predictor.py
+          - vllm_omni/model_executor/models/common/snake_activation.py
+          - vllm_omni/model_executor/models/qwen2_5_omni/qwen2_5_omni_thinker.py
+          - vllm_omni/model_executor/models/qwen3_tts/cuda_graph_decoder_wrapper.py
           - vllm_omni/model_executor/stage_input_processors/qwen3_omni.py
           - vllm_omni/deploy/qwen3_omni_moe.yaml
         timeout_in_minutes: 50
@@ -141,6 +141,11 @@ steps:
         source_file_dependencies:
           - tests/e2e/offline_inference/test_minicpmo_4_5.py
           - tests/e2e/online_serving/test_minicpmo_4_5.py
+          - vllm_omni/model_executor/models/minicpmo_4_5/
+          - vllm_omni/model_executor/models/step_audio2/
+          - vllm_omni/model_executor/models/cosyvoice3/code2wav_core/
+          - vllm_omni/model_executor/models/cosyvoice3/utils.py
+          - vllm_omni/model_executor/stage_input_processors/minicpmo_4_5_omni.py
           - vllm_omni/deploy/minicpmo_4_5.yaml
         timeout_in_minutes: 50
         commands:
@@ -162,70 +167,35 @@ steps:
           - tests/e2e/online_serving/helpers/minicpmo_realtime_duplex_scenarios.py
           - tests/e2e/online_serving/run_minicpmo_realtime_duplex_multi_session.py
           - tests/e2e/online_serving/run_minicpmo_realtime_duplex_soft_interrupt.py
+          - vllm_omni/deploy/minicpmo_4_5.yaml
           - tests/e2e/online_serving/helpers/minicpmo_4_5_duplex.py
           - vllm_omni/deploy/minicpmo_4_5_duplex.yaml
           - vllm_omni/experimental/fullduplex/
           - vllm_omni/model_executor/models/minicpmo_4_5/
+          - vllm_omni/model_executor/models/step_audio2/
+          - vllm_omni/model_executor/models/cosyvoice3/code2wav_core/
+          - vllm_omni/model_executor/models/cosyvoice3/utils.py
           - vllm_omni/model_executor/stage_input_processors/minicpmo_4_5_omni.py
           - vllm_omni/core/sched/
           - vllm_omni/outputs/multimodal_accumulation.py
           - vllm_omni/engine/stage_init_utils.py
           - vllm_omni/engine/async_omni_engine.py
+          - vllm_omni/engine/orchestrator.py
           - vllm_omni/entrypoints/cli/serve.py
+          - vllm_omni/entrypoints/openai/api_server.py
+          - vllm_omni/entrypoints/openai/duplex_capability.py
+          - vllm_omni/config/stage_config.py
         timeout_in_minutes: 50
         commands:
           - pytest -s -v tests/e2e/online_serving/test_minicpmo_4_5_duplex.py -m 'advanced_model and cuda' --run-level 'advanced_model'
         mirror_hardwares: h100_1
 
-      - label: "TTS · VoxCPM2 Test"
-        source_file_dependencies:
-          - tests/e2e/offline_inference/test_voxcpm2_tts.py
-          - tests/e2e/online_serving/test_voxcpm2_tts.py
-          - vllm_omni/model_executor/models/voxcpm2/
-          - vllm_omni/deploy/voxcpm2.yaml
-        timeout_in_minutes: 30
-        commands:
-          - |
-            timeout 30m bash -c "
-              export VLLM_LOGGING_LEVEL=DEBUG
-              export VLLM_WORKER_MULTIPROC_METHOD=spawn
-              pytest -s -v tests/e2e/offline_inference/test_voxcpm2_tts.py tests/e2e/online_serving/test_voxcpm2_tts.py -m 'advanced_model and cuda' --run-level 'advanced_model'
-            "
-        mirror_hardwares: l4_1
-
-      - label: "Diffusion · Qwen Image Edit Test"
-        source_file_dependencies:
-          - tests/e2e/online_serving/test_qwen_image_edit.py
-          - vllm_omni/diffusion/models/qwen_image/
-        timeout_in_minutes: 20
-        commands:
-          - pytest -s -v tests/e2e/online_serving/test_qwen_image_edit.py -m 'advanced_model and cuda' --run-level 'advanced_model'
-        mirror_hardwares: h100_1
-      - label: "Diffusion · Bagel Test"
-        source_file_dependencies:
-          - tests/e2e/offline_inference/test_bagel.py
-          - tests/e2e/online_serving/test_bagel.py
-          - vllm_omni/model_executor/models/bagel/
-          - vllm_omni/diffusion/models/bagel/
-          - vllm_omni/model_executor/stage_input_processors/bagel.py
-          - vllm_omni/deploy/bagel.yaml
-          - .buildkite/common/scripts/run_cov_split.sh
-          - pyproject.toml
-        # 20 -> 40: the coverage split runs the model twice.
-        timeout_in_minutes: 40
-        commands:
-          - |
-            .buildkite/common/scripts/run_cov_split.sh \
-              --model-id bagel \
-              --offline tests/e2e/offline_inference/test_bagel.py \
-              --online tests/e2e/online_serving/test_bagel.py \
-              -- -m 'advanced_model and cuda' --run-level 'advanced_model'
-        mirror_hardwares: h100_1
-
       - label: "Diffusion · Qwen Image Test"
         source_file_dependencies:
           - tests/e2e/online_serving/test_qwen_image.py
           - vllm_omni/diffusion/models/qwen_image/
+          - vllm_omni/diffusion/models/dmd2/
+          - vllm_omni/diffusion/distributed/autoencoders/autoencoder_kl_qwenimage.py
         commands:
           - |
             timeout 40m bash -c '
@@ -234,23 +204,12 @@ steps:
             '
         mirror_hardwares: h100_1
 
-      - label: "Diffusion · Qwen Image Layered Test"
-        source_file_dependencies:
-          - tests/e2e/online_serving/test_qwen_image_layered.py
-          - vllm_omni/diffusion/models/qwen_image/
-        commands:
-          - |
-            timeout 40m bash -c "
-              export VLLM_IMAGE_FETCH_TIMEOUT=60
-              pytest -s -v tests/e2e/online_serving/test_qwen_image_layered.py -m 'advanced_model and cuda' --run-level 'advanced_model'
-            "
-        mirror_hardwares: h100_1
-
       - label: "Diffusion · Wan22 Test"
         source_file_dependencies:
           - tests/e2e/offline_inference/test_wan22_t2v.py
           - tests/e2e/online_serving/test_wan22_t2v.py
           - vllm_omni/diffusion/models/wan2_2/
+          - vllm_omni/diffusion/models/dmd2/
         commands:
           - |
             timeout 40m bash -c "
@@ -258,21 +217,3 @@ steps:
               pytest -s -v tests/e2e/offline_inference/test_wan22_t2v.py tests/e2e/online_serving/test_wan22_t2v.py -m 'advanced_model and cuda and diffusion' --run-level 'advanced_model'
             "
         mirror_hardwares: h100_1
-
-      - label: "Diffusion · Z Image Test"
-        source_file_dependencies:
-          - tests/e2e/offline_inference/test_z_image.py
-          - tests/e2e/online_serving/test_z_image.py
-          - vllm_omni/diffusion/models/z_image/
-        timeout_in_minutes: 30
-        commands:
-          - pytest -s -v tests/e2e/offline_inference/test_z_image.py tests/e2e/online_serving/test_z_image.py -m 'advanced_model and cuda and diffusion' --run-level 'advanced_model'
-        mirror_hardwares: l4_1
-      - label: "Diffusion · FLUX.2 Klein Test"
-        source_file_dependencies:
-          - tests/e2e/offline_inference/test_flux2_klein.py
-          - vllm_omni/diffusion/models/flux2_klein/
-        timeout_in_minutes: 30
-        commands:
-          - pytest -s -v tests/e2e/offline_inference/test_flux2_klein.py -m 'advanced_model and cuda and diffusion' --run-level 'advanced_model'
-        mirror_hardwares: l4_1

.buildkite/cuda/test-nightly.yml

@@ -89,32 +89,6 @@ steps:
             exit $$EXIT
         mirror_hardwares: h100_2
 
-      - label: ":full_moon: Omni · Perf Test · vLLM Text"
-        key: nightly-omni-performance-vllm-text
-        timeout_in_minutes: 180
-        commands:
-          - export BENCHMARK_DIR=tests/dfx/perf/results
-          - |
-            set +e
-            pytest -s -v tests/dfx/perf/scripts/run_benchmark.py --test-config-file tests/dfx/perf/tests/test_qwen3_omni_vllm_text.json
-            EXIT=$$?
-            buildkite-agent artifact upload "tests/dfx/perf/results/*.json"
-            exit $$EXIT
-        mirror_hardwares: h100_2
-
-      - label: ":full_moon: Omni · Perf Test · Multi-Replica"
-        key: nightly-omni-performance-multi-replicas
-        timeout_in_minutes: 180
-        commands:
-          - export BENCHMARK_DIR=tests/dfx/perf/results
-          - |
-            set +e
-            pytest -s -v tests/dfx/perf/scripts/run_benchmark.py --test-config-file tests/dfx/perf/tests/test_qwen3_omni_multi_replicas.json
-            EXIT=$$?
-            buildkite-agent artifact upload "tests/dfx/perf/results/*.json"
-            exit $$EXIT
-        mirror_hardwares: h100_3
-
       - label: ":full_moon: Omni · MiniCPM-o 4.5 · Perf Test"
         key: nightly-omni-performance-minicpmo-4-5
         timeout_in_minutes: 180
@@ -159,19 +133,7 @@ steps:
           - export VLLM_ALLOW_LONG_MAX_MODEL_LEN="1"
           - export VLLM_USE_DEEP_GEMM="0"
           - export VLLM_MOE_USE_DEEP_GEMM="0"
-          - pytest -s -v tests/e2e/ -m "full_model and L4 and tts" --run-level "full_model" --ignore=tests/e2e/accuracy --ignore=tests/e2e/online_serving/test_voxcpm2_tts_expansion.py
-        mirror_hardwares: l4_1
-
-      - label: ":full_moon: TTS · Accuracy Test · VoxCPM2"
-        timeout_in_minutes: 30
-        commands:
-          - |
-            timeout 30m bash -c "
-              export VLLM_LOGGING_LEVEL=DEBUG
-              export VLLM_WORKER_MULTIPROC_METHOD=spawn
-              export VLLM_ALLOW_LONG_MAX_MODEL_LEN=1
-              pytest -s -v tests/e2e/online_serving/test_voxcpm2_tts_expansion.py -m 'full_model and cuda and tts' --run-level 'full_model'
-            "
+          - pytest -s -v tests/e2e/ -m "full_model and L4 and tts" --run-level "full_model" --ignore=tests/e2e/accuracy
         mirror_hardwares: l4_1
 
       - label: ":full_moon: TTS · Perf Test"
@@ -188,35 +150,6 @@ steps:
             exit $$EXIT
         mirror_hardwares: h100_2
 
-      - label: ":full_moon: TTS · VoxCPM2 · Perf Test"
-        key: nightly-tts-performance-voxcpm2
-        timeout_in_minutes: 180
-        commands:
-          - export BENCHMARK_DIR=tests/dfx/perf/results
-          - export VLLM_ALLOW_LONG_MAX_MODEL_LEN="1"
-          - |
-            set +e
-            pytest -s -v tests/dfx/perf/scripts/run_benchmark.py --test-config-file tests/dfx/perf/tests/test_voxcpm2.json
-            EXIT=$$?
-            buildkite-agent artifact upload "tests/dfx/perf/results/*.json"
-            exit $$EXIT
-        mirror_hardwares: h100_1
-
-      - label: ":full_moon: TTS · Higgs-Audio-V3 · Perf Test"
-        key: nightly-tts-performance-higgs-audio-v3
-        timeout_in_minutes: 180
-        commands:
-          - export BENCHMARK_DIR=tests/dfx/perf/results
-          - export VLLM_USE_DEEP_GEMM="0"
-          - export VLLM_MOE_USE_DEEP_GEMM="0"
-          - |
-            set +e
-            pytest -s -v tests/dfx/perf/scripts/run_benchmark.py --test-config-file tests/dfx/perf/tests/test_higgs_audio_v3.json
-            EXIT=$$?
-            buildkite-agent artifact upload "tests/dfx/perf/results/*.json"
-            exit $$EXIT
-        mirror_hardwares: h100_1
-
   - group: ":card_index_dividers: Tiny Model Tests [Multi-GPU]"
     key: nightly-tiny-model-test-group
     depends_on: upload-nightly-pipeline
@@ -246,8 +179,7 @@ steps:
       build.pull_request.labels includes "nightly-test" ||
       build.pull_request.labels includes "diffusion-x2iat-test"
     steps:
-      # Use explicit file shards so single-card cases do not reserve 2+ GPUs,
-      # and Bagel feature coverage is not duplicated via a broad tests/e2e sweep.
+      # Use explicit file shards so single-card cases do not reserve 2+ GPUs.
 
       - label: ":full_moon: Diffusion X2I(&A&T) · Function Test with H100 · Single-GPU"
         timeout_in_minutes: 120
@@ -255,15 +187,6 @@ steps:
           - >-
             pytest -sv
             tests/e2e/online_serving/test_qwen_image_expansion.py
-            tests/e2e/online_serving/test_qwen_image_edit_expansion.py
-            tests/e2e/online_serving/test_qwen_image_layered_expansion.py
-            tests/e2e/online_serving/test_boogu_image.py
-            tests/e2e/online_serving/test_boogu_image_edit.py
-            tests/e2e/online_serving/test_boogu_image_expansion.py
-            tests/e2e/online_serving/test_boogu_image_edit_expansion.py
-            tests/e2e/online_serving/test_bagel_expansion.py
-            tests/e2e/online_serving/test_krea2_expansion.py
-            tests/e2e/offline_inference/test_krea2_expansion.py
             -m "full_model and diffusion and H100 and not distributed_cuda"
             --run-level "full_model"
         mirror_hardwares: h100_1
@@ -274,47 +197,14 @@ steps:
           - >-
             pytest -sv
             tests/e2e/online_serving/test_qwen_image_expansion.py
-            tests/e2e/online_serving/test_qwen_image_edit_expansion.py
-            tests/e2e/offline_inference/test_krea2_expansion.py
             -m "full_model and diffusion and H100 and distributed_cuda"
             --run-level "full_model"
         mirror_hardwares: h100_2
 
-      - label: ":full_moon: Diffusion X2I(&A&T) · Function Test with H100 · Multi-GPU Layered/LongCat/FLUX"
-        timeout_in_minutes: 90
-        commands:
-          - >-
-            pytest -sv
-            tests/e2e/online_serving/test_qwen_image_layered_expansion.py
-            -m "full_model and diffusion and H100 and distributed_cuda"
-            --run-level "full_model"
-        mirror_hardwares: h100_2
-
-      - label: ":full_moon: Diffusion X2I(&A&T) · Function Test with H100 · Multi-GPU Bagel"
-        timeout_in_minutes: 120
-        commands:
-          - >-
-            pytest -sv
-            tests/e2e/online_serving/test_bagel_expansion.py
-            -m "full_model and diffusion and H100 and distributed_cuda"
-            --run-level "full_model"
-        mirror_hardwares: h100_2
-
-      - label: ":full_moon: Diffusion X2I(&A&T) · GLM-Image Function Test with H100"
-        timeout_in_minutes: 120
-        commands:
-          - >-
-            pytest -sv
-            tests/e2e/online_serving/test_glm_image_expansion.py
-            tests/e2e/offline_inference/test_glm_image_autoround_w4a16_expansion.py
-            -m "full_model and diffusion and H100"
-            --run-level "full_model"
-        mirror_hardwares: h100_2
-
       - label: ":full_moon: Diffusion X2I(&A&T) · Function Test with L4"
         timeout_in_minutes: 120
         commands:
-          - pytest -sv tests/e2e/ -k "not test_wan and not test_bagel_expansion and not hunyuan" -m "full_model and diffusion and L4" --run-level "full_model" --ignore=tests/e2e/accuracy
+          - pytest -sv tests/e2e/ -k "not test_wan and not hunyuan" -m "full_model and diffusion and L4" --run-level "full_model" --ignore=tests/e2e/accuracy
         mirror_hardwares: l4_4
 
       - label: ":full_moon: Diffusion X2I(&A&T) · Doc Test"
@@ -327,7 +217,7 @@ steps:
         timeout_in_minutes: 180
         commands:
           - export VLLM_HTTP_TIMEOUT_KEEP_ALIVE=120
-          - pytest -s -v tests/e2e/accuracy/test_qwen_image*.py --run-level full_model
+          - pytest -s -v tests/e2e/accuracy/test_qwen_image.py --run-level full_model
           - pytest -s -v tests/e2e/accuracy/test_diffusers_backend_similarity.py -k '2i_matches_diffusers' -m full_model --run-level full_model
         mirror_hardwares: h100_1
 
@@ -339,6 +229,12 @@ steps:
           - pytest -s -v tests/e2e/accuracy/test_hunyuan_image3_pixel_accuracy.py -m full_model --run-level full_model
         mirror_hardwares: h100_4
 
+      - label: ":full_moon: HiDream-O1-Image · Accuracy Test"
+        timeout_in_minutes: 180
+        commands:
+          - pytest -s -v tests/e2e/accuracy/test_hidream_o1_image.py -m full_model --run-level full_model
+        mirror_hardwares: h100_2
+
       - label: ":full_moon: Diffusion X2I(&A&T) · Perf Test · Qwen-Image"
         key: nightly-diffusion-x2iat-performance-qwen-image
         timeout_in_minutes: 180
@@ -355,70 +251,6 @@ steps:
             exit $$EXIT
         mirror_hardwares: h100_4
 
-      - label: ":full_moon: Diffusion X2I(&A&T) · Perf Test · Qwen-Image-Edit"
-        key: nightly-diffusion-x2iat-performance-qwen-image-edit
-        timeout_in_minutes: 180
-        commands:
-          - export DIFFUSION_BENCHMARK_DIR=tests/dfx/perf/results
-          - export DIFFUSION_ATTENTION_BACKEND=FLASH_ATTN
-          - export CACHE_DIT_VERSION=1.5.0
-          - |
-            set +e
-            pytest -s -v tests/dfx/perf/scripts/run_diffusion_benchmark.py --test-config-file tests/dfx/perf/tests/test_qwen_image_edit_vllm_omni.json
-            EXIT=$$?
-            buildkite-agent artifact upload "tests/dfx/perf/results/diffusion_result_*.json"
-            buildkite-agent artifact upload "tests/dfx/perf/results/logs/*.log"
-            exit $$EXIT
-        mirror_hardwares: h100_4
-
-      - label: ":full_moon: Diffusion X2I(&A&T) · Perf Test · Qwen-Image-Edit-2511"
-        key: nightly-diffusion-x2iat-performance-qwen-image-edit-2511
-        timeout_in_minutes: 180
-        commands:
-          - export DIFFUSION_BENCHMARK_DIR=tests/dfx/perf/results
-          - export DIFFUSION_ATTENTION_BACKEND=FLASH_ATTN
-          - export CACHE_DIT_VERSION=1.5.0
-          - |
-            set +e
-            pytest -s -v tests/dfx/perf/scripts/run_diffusion_benchmark.py --test-config-file tests/dfx/perf/tests/test_qwen_image_edit_2511_vllm_omni.json
-            EXIT=$$?
-            buildkite-agent artifact upload "tests/dfx/perf/results/diffusion_result_*.json"
-            buildkite-agent artifact upload "tests/dfx/perf/results/logs/*.log"
-            exit $$EXIT
-        mirror_hardwares: h100_4
-
-      - label: ":full_moon: Diffusion X2I(&A&T) · Perf Test · Qwen-Image-Layered"
-        key: nightly-diffusion-x2iat-performance-qwen-image-layered
-        timeout_in_minutes: 120
-        commands:
-          - export DIFFUSION_BENCHMARK_DIR=tests/dfx/perf/results
-          - export DIFFUSION_ATTENTION_BACKEND=FLASH_ATTN
-          - export CACHE_DIT_VERSION=1.5.0
-          - |
-            set +e
-            pytest -s -v tests/dfx/perf/scripts/run_diffusion_benchmark.py --test-config-file tests/dfx/perf/tests/test_qwen_image_layered_vllm_omni.json
-            EXIT=$$?
-            buildkite-agent artifact upload "tests/dfx/perf/results/diffusion_result_*.json"
-            buildkite-agent artifact upload "tests/dfx/perf/results/logs/*.log"
-            exit $$EXIT
-        mirror_hardwares: h100_1
-
-      - label: ":full_moon: Diffusion X2I(&A&T) · Perf Test · BAGEL"
-        key: nightly-diffusion-x2iat-performance-bagel
-        timeout_in_minutes: 180
-        commands:
-          - export DIFFUSION_BENCHMARK_DIR=tests/dfx/perf/results
-          - export DIFFUSION_ATTENTION_BACKEND=FLASH_ATTN
-          - export CACHE_DIT_VERSION=1.5.0
-          - |
-            set +e
-            pytest -s -v tests/dfx/perf/scripts/run_diffusion_benchmark.py --test-config-file tests/dfx/perf/tests/test_bagel_vllm_omni.json
-            EXIT=$$?
-            buildkite-agent artifact upload "tests/dfx/perf/results/diffusion_result_*.json"
-            buildkite-agent artifact upload "tests/dfx/perf/results/logs/*.log"
-            exit $$EXIT
-        mirror_hardwares: h100_1
-
   - group: ":card_index_dividers: Diffusion X2V Model Test"
     key: nightly-diffusion-x2v-group
     depends_on: upload-nightly-pipeline
@@ -440,12 +272,6 @@ steps:
           - pytest -s -v tests/e2e/online_serving/test_wan22_expansion.py -k "i2v" -m "full_model and cuda" --run-level "full_model"
         mirror_hardwares: h100_2
 
-      - label: ":full_moon: Diffusion X2V · LingBot Video MoE Function Test"
-        timeout_in_minutes: 90
-        commands:
-          - pytest -s -v tests/e2e/online_serving/test_lingbot_video_moe.py -k "video_generation" -m "full_model and cuda" --run-level "full_model"
-        mirror_hardwares: h100_1
-
       - label: ":full_moon: Diffusion X2V · Doc Test"
         timeout_in_minutes: 60
         commands:
@@ -456,7 +282,6 @@ steps:
         timeout_in_minutes: 90
         commands:
           - pytest -s -v tests/e2e/offline_inference/test_wan22_autoround_w4a16_expansion.py -m "full_model and cuda" --run-level "full_model"
-          - pytest -s -v tests/e2e/online_serving/test_wan_2_1_vace_expansion.py tests/e2e/online_serving/test_ltx.py -m "full_model and cuda" --run-level "full_model"
         mirror_hardwares: h100_2
 
       - label: ":full_moon: Diffusion X2V · Wan2.2 Accuracy Test"
@@ -477,16 +302,8 @@ steps:
         commands:
           - pytest -s -v tests/e2e/accuracy/hunyuanvideo15_t2v/test_hunyuanvideo15_t2v_video_similarity.py -m full_model --run-level full_model
           - pytest -s -v tests/e2e/accuracy/hunyuanvideo15_i2v/test_hunyuanvideo15_i2v_video_similarity.py -m full_model --run-level full_model
-          - pytest -s -v tests/e2e/accuracy/test_video_streaming_output_similarity.py -m full_model --run-level full_model
         mirror_hardwares: h100_2
 
-      - label: ":full_moon: Diffusion X2V · LTX Accuracy Test"
-        timeout_in_minutes: 180
-        commands:
-          - pytest -s -v 'tests/e2e/accuracy/ltx/test_ltx_official_similarity.py::test_ltx_matches_official[ltx2_3_i2v]' -m full_model --run-level full_model
-          - pytest -s -v 'tests/e2e/accuracy/ltx/test_ltx_official_similarity.py::test_ltx_matches_official[ltx23_two_stage_layer_fused_i2v]' -m full_model --run-level full_model
-        mirror_hardwares: h100_1
-
       - label: ":full_moon: Diffusion X2V · Perf Test"
         key: nightly-diffusion-x2v-performance
         timeout_in_minutes: 180
@@ -562,16 +379,6 @@ steps:
           - pytest -sv tests/diffusion/quantization -m 'full_model and cuda and L4' --run-level "full_model"
         mirror_hardwares: l4_1
 
-  - label: ":bar_chart: Testcase Statistics"
-    key: nightly-testcase-statistics
-    timeout_in_minutes: 120
-    depends_on: upload-nightly-pipeline
-    if: build.env("NIGHTLY") == "1"
-    commands:
-      - python tools/nightly/buildkite_testcase_statistics.py -o tests/dfx/perf/results/buildkite_testcase_statistics.html
-      - buildkite-agent artifact upload "tests/dfx/perf/results/*.html"
-    mirror_hardwares: h100_1
-
   # NOTE: ":email: Nightly Collection & Email" is deprecated and will be removed soon.
   # Do not add new depends_on entries or extend this step.
   - label: ":email: Nightly Collection & Email"
@@ -579,35 +386,19 @@ steps:
     depends_on:
       - nightly-omni-performance-no-async-chunk
       - nightly-omni-performance-async-chunk
-      - nightly-omni-performance-vllm-text
-      - nightly-omni-performance-multi-replicas
       - nightly-tts-performance
-      - nightly-tts-performance-higgs-audio-v3
       - nightly-diffusion-x2iat-performance-qwen-image
-      - nightly-diffusion-x2iat-performance-qwen-image-edit
-      - nightly-diffusion-x2iat-performance-qwen-image-edit-2511
-      - nightly-diffusion-x2iat-performance-qwen-image-layered
-      - nightly-diffusion-x2iat-performance-bagel
       - nightly-diffusion-x2v-performance
-      - nightly-testcase-statistics
     if: build.env("NIGHTLY") == "1" && build.env("EMAIL_DISTRIBUTION") == "1"
     commands:
       - pip install openpyxl
       - export DEFAULT_INPUT_DIR=tests/dfx/perf/results
       - export DEFAULT_OUTPUT_DIR=tests/dfx/perf/results
       - buildkite-agent artifact download "tests/dfx/perf/results/*.json" . --step nightly-tts-performance
-      - buildkite-agent artifact download "tests/dfx/perf/results/*.json" . --step nightly-tts-performance-higgs-audio-v3
       - buildkite-agent artifact download "tests/dfx/perf/results/*.json" . --step nightly-omni-performance-no-async-chunk
       - buildkite-agent artifact download "tests/dfx/perf/results/*.json" . --step nightly-omni-performance-async-chunk
-      - buildkite-agent artifact download "tests/dfx/perf/results/*.json" . --step nightly-omni-performance-vllm-text
-      - buildkite-agent artifact download "tests/dfx/perf/results/*.json" . --step nightly-omni-performance-multi-replicas
       - buildkite-agent artifact download "tests/dfx/perf/results/*.json" . --step nightly-diffusion-x2iat-performance-qwen-image
-      - buildkite-agent artifact download "tests/dfx/perf/results/*.json" . --step nightly-diffusion-x2iat-performance-qwen-image-edit
-      - buildkite-agent artifact download "tests/dfx/perf/results/*.json" . --step nightly-diffusion-x2iat-performance-qwen-image-edit-2511
-      - buildkite-agent artifact download "tests/dfx/perf/results/*.json" . --step nightly-diffusion-x2iat-performance-qwen-image-layered
-      - buildkite-agent artifact download "tests/dfx/perf/results/*.json" . --step nightly-diffusion-x2iat-performance-bagel
       - buildkite-agent artifact download "tests/dfx/perf/results/*.json" . --step nightly-diffusion-x2v-performance
-      - buildkite-agent artifact download "tests/dfx/perf/results/*.html" . --step nightly-testcase-statistics
       - python tools/nightly/generate_nightly_perf_excel.py
       - python tools/nightly/generate_nightly_perf_html.py
       - python tools/nightly/send_nightly_email.py --report-file "tests/dfx/perf/results/*.xlsx, tests/dfx/perf/results/*.html"

.buildkite/cuda/test-ready.yml

@@ -9,23 +9,27 @@ steps:
     depends_on: upload-ready-pipeline
     steps:
       - label: "Simple · Model Executor Test"
+        timeout_in_minutes: 20
         commands:
-          - "timeout 20m pytest -sv tests/model_executor -m 'core_model and cpu' --cov=vllm_omni --cov-report=term-missing:skip-covered --cov-report=xml"
+          - "pytest -sv tests/model_executor -m 'core_model and cpu'"
         mirror_hardwares: l4_1
 
       - label: "Simple · Diffusion Test"
+        timeout_in_minutes: 20
         commands:
-          - "timeout 20m pytest -sv tests/diffusion -m 'core_model and cpu' --cov=vllm_omni --cov-report=term-missing:skip-covered --cov-report=xml"
+          - "pytest -sv tests/diffusion -m 'core_model and cpu'"
         mirror_hardwares: l4_1
 
       - label: "Simple · Engine&Entrypoints Test"
+        timeout_in_minutes: 20
         commands:
-          - "timeout 20m pytest -sv tests/entrypoints tests/engine -m 'core_model and cpu' --cov=vllm_omni --cov-report=term-missing:skip-covered --cov-report=xml"
+          - "pytest -sv tests/entrypoints tests/engine -m 'core_model and cpu'"
         mirror_hardwares: l4_1
 
       - label: "Simple · Other Test"
+        timeout_in_minutes: 20
         commands:
-          - "timeout 20m pytest -sv tests/ -m 'core_model and cpu' --ignore=tests/diffusion --ignore=tests/model_executor --ignore=tests/entrypoints --ignore=tests/engine --cov=vllm_omni --cov-report=term-missing:skip-covered --cov-report=xml"
+          - "pytest -sv tests/ -m 'core_model and cpu' --ignore=tests/diffusion --ignore=tests/model_executor --ignore=tests/entrypoints --ignore=tests/engine"
         mirror_hardwares: l4_1
 
   - group: ":card_index_dividers: Diffusion Test"
@@ -76,7 +80,7 @@ steps:
 
       - label: "Tiny Model Tests · Diffusion (base)"
         commands:
-          - timeout 15m pytest -sv tests/model_tests/diffusion -m 'core_model and cuda' --run-level "core_model"
+          - timeout 15m pytest -sv -n 4 tests/model_tests/diffusion -m 'core_model and cuda' --run-level "core_model"
         mirror_hardwares: l4_1
 
   - label: "Engine&Model Executor Test"
@@ -121,32 +125,8 @@ steps:
           - vllm_omni/outputs/
         commands:
           - timeout 20m pytest -s -v tests/e2e/features/custom_pipeline/ -m "core_model"
-        agents:
-          queue: "gpu_1_queue"
-        plugins:
-          - docker#v5.2.0:
-              image: public.ecr.aws/q9t5s3a7/vllm-ci-test-repo:$BUILDKITE_COMMIT
-              always-pull: true
-              propagate-environment: true
-              # vLLM 0.27 added a pre-flight capacity check in
-              # distributed/device_communicators/shm_broadcast.py. Without this
-              # the container gets Docker's 64 MiB default and every test in
-              # this step dies at engine start with
-              #   RuntimeError: Insufficient space in /dev/shm:
-              #                 240 MiB required, 64 MiB available
-              # (9 failed in 8.7s on build #2949). vLLM 0.26 had no such check
-              # and allocated regardless, so this step passed on main while
-              # being under-provisioned the whole time — 0.27 turned a silent
-              # risk into an explicit error.
-              #
-              # This is the only inline docker plugin block in .buildkite/cuda/;
-              # every other step inherits common/ci_mirror_hardwares.yml, which
-              # already sets 8gb. Matching that rather than inventing a value.
-              shm-size: "8gb"
-              environment:
-                - "HF_HOME=/fsx/hf_cache"
-              volumes:
-                - "/fsx/hf_cache:/fsx/hf_cache"
+        mirror_hardwares: l4_1
+
 
       - label: "RLHF · VeRL-Omni E2E Test"
         source_file_dependencies:
@@ -168,7 +148,13 @@ steps:
       - label: "Omni · Qwen3-Omni Test"
         source_file_dependencies:
           - tests/e2e/online_serving/test_qwen3_omni.py
+          - tests/helpers/runtime.py
+          - vllm_omni/entrypoints/openai/serving_chat.py
           - vllm_omni/model_executor/models/qwen3_omni/
+          - vllm_omni/model_executor/models/common/qwen3_code_predictor.py
+          - vllm_omni/model_executor/models/common/snake_activation.py
+          - vllm_omni/model_executor/models/qwen2_5_omni/qwen2_5_omni_thinker.py
+          - vllm_omni/model_executor/models/qwen3_tts/cuda_graph_decoder_wrapper.py
           - vllm_omni/model_executor/stage_input_processors/qwen3_omni.py
           - vllm_omni/deploy/qwen3_omni_moe.yaml
         commands:
@@ -181,6 +167,11 @@ steps:
       - label: "Omni · MiniCPM-o 4.5 Offline Test"
         source_file_dependencies:
           - tests/e2e/offline_inference/test_minicpmo_4_5.py
+          - vllm_omni/model_executor/models/minicpmo_4_5/
+          - vllm_omni/model_executor/models/step_audio2/
+          - vllm_omni/model_executor/models/cosyvoice3/code2wav_core/
+          - vllm_omni/model_executor/models/cosyvoice3/utils.py
+          - vllm_omni/model_executor/stage_input_processors/minicpmo_4_5_omni.py
           - vllm_omni/deploy/minicpmo_4_5.yaml
         commands:
           - |
@@ -192,6 +183,11 @@ steps:
       - label: "Omni · MiniCPM-o 4.5 Online Test"
         source_file_dependencies:
           - tests/e2e/online_serving/test_minicpmo_4_5.py
+          - vllm_omni/model_executor/models/minicpmo_4_5/
+          - vllm_omni/model_executor/models/step_audio2/
+          - vllm_omni/model_executor/models/cosyvoice3/code2wav_core/
+          - vllm_omni/model_executor/models/cosyvoice3/utils.py
+          - vllm_omni/model_executor/stage_input_processors/minicpmo_4_5_omni.py
           - vllm_omni/deploy/minicpmo_4_5.yaml
         commands:
           - |
@@ -210,16 +206,24 @@ steps:
           - tests/e2e/online_serving/helpers/minicpmo_realtime_duplex_scenarios.py
           - tests/e2e/online_serving/run_minicpmo_realtime_duplex_multi_session.py
           - tests/e2e/online_serving/run_minicpmo_realtime_duplex_soft_interrupt.py
+          - vllm_omni/deploy/minicpmo_4_5.yaml
           - tests/e2e/online_serving/helpers/minicpmo_4_5_duplex.py
           - vllm_omni/deploy/minicpmo_4_5_duplex.yaml
           - vllm_omni/experimental/fullduplex/
           - vllm_omni/model_executor/models/minicpmo_4_5/
+          - vllm_omni/model_executor/models/step_audio2/
+          - vllm_omni/model_executor/models/cosyvoice3/code2wav_core/
+          - vllm_omni/model_executor/models/cosyvoice3/utils.py
           - vllm_omni/model_executor/stage_input_processors/minicpmo_4_5_omni.py
           - vllm_omni/core/sched/
           - vllm_omni/outputs/multimodal_accumulation.py
           - vllm_omni/engine/stage_init_utils.py
           - vllm_omni/engine/async_omni_engine.py
+          - vllm_omni/engine/orchestrator.py
           - vllm_omni/entrypoints/cli/serve.py
+          - vllm_omni/entrypoints/openai/api_server.py
+          - vllm_omni/entrypoints/openai/duplex_capability.py
+          - vllm_omni/config/stage_config.py
         commands:
           - |
             timeout 20m bash -c "
@@ -239,25 +243,13 @@ steps:
           - pytest -s -v tests/dfx/perf/scripts/run_benchmark.py --test-config-file tests/dfx/perf/tests/test_minicpmo_4_5_ready.json -m 'core_model and cuda'
         mirror_hardwares: h100_1
 
-      - label: "TTS · VoxCPM2 Test"
-        source_file_dependencies:
-          - tests/e2e/online_serving/test_voxcpm2_tts.py
-          - vllm_omni/model_executor/models/voxcpm2/
-          - vllm_omni/deploy/voxcpm2.yaml
-        timeout_in_minutes: 30
-        commands:
-          - |
-            timeout 30m bash -c "
-              export VLLM_LOGGING_LEVEL=DEBUG
-              export VLLM_WORKER_MULTIPROC_METHOD=spawn
-              pytest -s -v tests/e2e/online_serving/test_voxcpm2_tts.py -m 'core_model' --run-level 'core_model'
-            "
-        mirror_hardwares: l4_1
-
       - label: "TTS · Qwen3-TTS CustomVoice Test"
         source_file_dependencies:
           - tests/e2e/online_serving/test_qwen3_tts_customvoice.py
           - vllm_omni/model_executor/models/qwen3_tts/
+          - vllm_omni/model_executor/models/common/qwen3_code_predictor.py
+          - vllm_omni/model_executor/models/common/snake_activation.py
+          - vllm_omni/model_executor/models/common/alias_free_activation.py
           - vllm_omni/model_executor/stage_input_processors/qwen3_tts.py
           - vllm_omni/deploy/qwen3_tts.yaml
         commands:
@@ -272,6 +264,9 @@ steps:
         source_file_dependencies:
           - tests/e2e/online_serving/test_qwen3_tts_base.py
           - vllm_omni/model_executor/models/qwen3_tts/
+          - vllm_omni/model_executor/models/common/qwen3_code_predictor.py
+          - vllm_omni/model_executor/models/common/snake_activation.py
+          - vllm_omni/model_executor/models/common/alias_free_activation.py
           - vllm_omni/model_executor/stage_input_processors/qwen3_tts.py
           - vllm_omni/deploy/qwen3_tts.yaml
         commands:
@@ -284,52 +279,11 @@ steps:
             "
         mirror_hardwares: l4_1
 
-      - label: "TTS · Higgs-Audio-V3 Test"
-        source_file_dependencies:
-          - tests/e2e/online_serving/test_higgs_audio_v3.py
-          - vllm_omni/model_executor/models/higgs_audio_v3/
-          - vllm_omni/model_executor/stage_input_processors/higgs_audio_v3.py
-          - vllm_omni/deploy/higgs_multimodal_qwen3.yaml
-        commands:
-          - |
-            timeout 20m bash -c "
-              export VLLM_WORKER_MULTIPROC_METHOD=spawn
-              export VLLM_USE_DEEP_GEMM=0
-              export VLLM_MOE_USE_DEEP_GEMM=0
-              pytest -s -v tests/e2e/online_serving/test_higgs_audio_v3.py -m 'core_model' --run-level 'core_model'
-            "
-        mirror_hardwares: l4_1
-
-      - label: "Diffusion · Qwen Image Edit Test"
-        source_file_dependencies:
-          - tests/e2e/online_serving/test_qwen_image_edit.py
-          - vllm_omni/diffusion/models/qwen_image/
-        commands:
-          - |
-            timeout 20m bash -c "
-              pytest -s -v tests/e2e/online_serving/test_qwen_image_edit.py -m 'core_model' --run-level 'core_model'
-            "
-        mirror_hardwares: h100_1
-
-      - label: "Diffusion · Bagel Test"
-        source_file_dependencies:
-          - tests/e2e/online_serving/test_bagel.py
-          - vllm_omni/model_executor/models/bagel/
-          - vllm_omni/diffusion/models/bagel/
-          - vllm_omni/model_executor/stage_input_processors/bagel.py
-          - vllm_omni/deploy/bagel.yaml
-        commands:
-          - |
-            timeout 40m bash -c "
-              export VLLM_IMAGE_FETCH_TIMEOUT=60
-              pytest -s -v tests/e2e/online_serving/test_bagel.py -m 'core_model' --run-level 'core_model'
-            "
-        mirror_hardwares: h100_1
-
       - label: "Diffusion · Wan22 Test"
         source_file_dependencies:
           - tests/e2e/online_serving/test_wan22_t2v.py
           - vllm_omni/diffusion/models/wan2_2/
+          - vllm_omni/diffusion/models/dmd2/
         commands:
           - |
             timeout 40m bash -c "
@@ -342,6 +296,9 @@ steps:
         source_file_dependencies:
           - tests/e2e/online_serving/test_cosmos3.py
           - vllm_omni/diffusion/models/cosmos3/
+          - vllm_omni/diffusion/distributed/autoencoders/autoencoder_kl_wan.py
+          - vllm_omni/experimental/world_models/
+          - vllm_omni/model_extras/cosmos3.py
         commands:
           - |
             timeout 40m bash -c "
@@ -354,6 +311,8 @@ steps:
         source_file_dependencies:
           - tests/e2e/online_serving/test_qwen_image.py
           - vllm_omni/diffusion/models/qwen_image/
+          - vllm_omni/diffusion/models/dmd2/
+          - vllm_omni/diffusion/distributed/autoencoders/autoencoder_kl_qwenimage.py
         commands:
           - |
             timeout 40m bash -c "
@@ -361,15 +320,3 @@ steps:
               pytest -s -v tests/e2e/online_serving/test_qwen_image.py -m 'core_model' --run-level 'core_model'
             "
         mirror_hardwares: h100_1
-
-      - label: "Diffusion · Qwen Image Layered Test"
-        source_file_dependencies:
-          - tests/e2e/online_serving/test_qwen_image_layered.py
-          - vllm_omni/diffusion/models/qwen_image/
-        commands:
-          - |
-            timeout 40m bash -c "
-              export VLLM_IMAGE_FETCH_TIMEOUT=60
-              pytest -s -v tests/e2e/online_serving/test_qwen_image_layered.py -m 'core_model' --run-level 'core_model'
-            "
-        mirror_hardwares: h100_1

.buildkite/cuda/test-weekly.yml

@@ -5,59 +5,91 @@ env:
   HF_HUB_ETAG_TIMEOUT: 60
 
 steps:
-
-  - label: "Reliability Test - qwen3-omni"
-    timeout_in_minutes: 180
+  - label: "Simple · CPU Coverage Test"
+    key: weekly-simple-cpu-coverage
     depends_on: upload-weekly-pipeline
-    if: build.env("WEEKLY") == "1" || build.pull_request.labels includes "weekly-test"
+    if: build.env("WEEKLY") == "1"
+    timeout_in_minutes: 60
     commands:
-      - pytest -s -v tests/dfx/reliability/test_reliability_qwen3_omni.py -m "slow"
-    mirror_hardwares: h100_2
-
-  - label: "Reliability Test - wan22"
-    timeout_in_minutes: 180
-    depends_on: upload-weekly-pipeline
-    if: build.env("WEEKLY") == "1" || build.pull_request.labels includes "weekly-test"
-    commands:
-      - pytest -s -v tests/dfx/reliability/test_reliability_wan22.py -m "slow"
-    mirror_hardwares: h100_2
-
-  - label: "Reliability Test - hunyuan-image"
-    timeout_in_minutes: 180
-    depends_on: upload-weekly-pipeline
-    if: build.env("WEEKLY") == "1" || build.pull_request.labels includes "weekly-test"
-    commands:
-      - pytest -s -v tests/dfx/reliability/test_reliability_hunyuan_image.py -m "slow"
-    mirror_hardwares: h100_4
-
-  - label: "Reliability Test - voxcpm2"
-    timeout_in_minutes: 180
-    depends_on: upload-weekly-pipeline
-    if: build.env("WEEKLY") == "1" || build.pull_request.labels includes "weekly-test"
-    commands:
-      - pytest -s -v tests/dfx/reliability/test_reliability_voxcpm2.py -m "slow"
+      - |
+        set +e
+        REPORT="coverage-core-model-cpu-$${BUILDKITE_STEP_ID:-local}.xml"
+        pytest -sv tests/ -m 'core_model and cpu' --cov=vllm_omni --cov-report=term-missing:skip-covered --cov-report=xml:$${REPORT}
+        EXIT=$$?
+        gzip -9 -f "$${REPORT}"
+        buildkite-agent artifact upload "$${REPORT}.gz"
+        exit $$EXIT
     mirror_hardwares: l4_1
 
-  - group: "Reliability Test - Invalid parameters Test"
+  - group: "Reliability Test"
     depends_on: upload-weekly-pipeline
     if: build.env("WEEKLY") == "1" || build.pull_request.labels includes "weekly-test"
     steps:
 
-      - label: "Invalid parameters Test · H100"
+      - label: "Reliability Test · qwen3-omni"
+        timeout_in_minutes: 180
+        commands:
+          - pytest -s -v tests/dfx/reliability/test_reliability_qwen3_omni.py -m "slow"
+        mirror_hardwares: h100_2
+
+      - label: "Reliability Test · wan22"
+        timeout_in_minutes: 180
+        commands:
+          - pytest -s -v tests/dfx/reliability/test_reliability_wan22.py -m "slow"
+        mirror_hardwares: h100_2
+
+      - label: "Reliability Test · hunyuan-image"
+        timeout_in_minutes: 180
+        commands:
+          - pytest -s -v tests/dfx/reliability/test_reliability_hunyuan_image.py -m "slow"
+        mirror_hardwares: h100_4
+
+      - label: "Reliability Test · Invalid parameters · H100"
         timeout_in_minutes: 120
         commands:
           - pytest -s -v tests/dfx/reliability/invalid_param_test/ -m "slow and H100"
         mirror_hardwares: h100_2
 
-      - label: "Invalid parameters Test · L4"
+      - label: "Reliability Test · Invalid parameters · L4"
         timeout_in_minutes: 120
         commands:
           - pytest -s -v tests/dfx/reliability/invalid_param_test/ -m "slow and L4"
         mirror_hardwares: l4_1
 
-  - group: ":card_index_dividers: E2E Tests"
+  - group: "Perf Test"
     depends_on: upload-weekly-pipeline
     if: build.env("WEEKLY") == "1" || build.pull_request.labels includes "weekly-test"
+    steps:
+
+      - label: ":full_moon: Omni · Perf Test · vLLM Text"
+        key: weekly-omni-performance-vllm-text
+        timeout_in_minutes: 180
+        commands:
+          - export BENCHMARK_DIR=tests/dfx/perf/results
+          - |
+            set +e
+            pytest -s -v tests/dfx/perf/scripts/run_benchmark.py --test-config-file tests/dfx/perf/tests/test_qwen3_omni_vllm_text.json
+            EXIT=$$?
+            buildkite-agent artifact upload "tests/dfx/perf/results/*.json"
+            exit $$EXIT
+        mirror_hardwares: h100_2
+
+      - label: ":full_moon: Omni · Perf Test · Multi-Replica"
+        key: weekly-omni-performance-multi-replicas
+        timeout_in_minutes: 180
+        commands:
+          - export BENCHMARK_DIR=tests/dfx/perf/results
+          - |
+            set +e
+            pytest -s -v tests/dfx/perf/scripts/run_benchmark.py --test-config-file tests/dfx/perf/tests/test_qwen3_omni_multi_replicas.json
+            EXIT=$$?
+            buildkite-agent artifact upload "tests/dfx/perf/results/*.json"
+            exit $$EXIT
+        mirror_hardwares: h100_3
+
+  - group: ":card_index_dividers: E2E Tests"
+    depends_on: upload-weekly-pipeline
+    if: build.env("NON_CRITICAL") == "1"
     steps:
 
       - label: "Omni · L4"
@@ -83,7 +115,7 @@ steps:
         mirror_hardwares: h100_2
 
       - label: "TTS · L4"
-        timeout_in_minutes: 90
+        timeout_in_minutes: 120
         commands:
           - export VLLM_ALLOW_LONG_MAX_MODEL_LEN="1"
           - export VLLM_USE_DEEP_GEMM="0"
@@ -108,6 +140,22 @@ steps:
             --run-level "full_model"
         mirror_hardwares: h100_1
 
+      # Own step: the image has no NanoCodec, so this file import-skips out of
+      # the mark-selected TTS steps above. The overrides keep nemo_toolkit[tts]
+      # from downgrading transformers; see
+      # examples/offline_inference/text_to_speech/README.md.
+      - label: "TTS · Gepard-1.0 · L4"
+        timeout_in_minutes: 60
+        commands:
+          - export VLLM_ALLOW_LONG_MAX_MODEL_LEN="1"
+          - printf '%s\n' "transformers>=5.5.3" "huggingface-hub>=1.0" > /tmp/gepard-overrides.txt
+          - uv pip install --python "$(python3 -c 'import sys; print(sys.executable)')" "nemo_toolkit[tts]==2.7.3" --overrides /tmp/gepard-overrides.txt
+          - >-
+            pytest -s -v tests/e2e/offline_inference/test_gepard_tts.py
+            -m "tts and cuda"
+            --run-level "full_model"
+        mirror_hardwares: l4_1
+
       - label: "Diffusion · L4"
         timeout_in_minutes: 60
         commands:
@@ -118,7 +166,7 @@ steps:
         mirror_hardwares: l4_4
 
       - label: "Diffusion · H100 · Single-GPU"
-        timeout_in_minutes: 120
+        timeout_in_minutes: 150
         commands:
           - >-
             pytest -sv
@@ -133,4 +181,4 @@ steps:
             pytest -sv
             tests/e2e/ -m "slow and diffusion and H100 and distributed_cuda"
             --run-level "full_model"
-        mirror_hardwares: h100_2
+        mirror_hardwares: h100_4

.buildkite/npu/test-npu-nightly.yml

@@ -69,12 +69,67 @@ steps:
           VLLM_ALLOW_LONG_MAX_MODEL_LEN: "1"
           HF_TOKEN: "${HF_TOKEN}"
         commands:
+          # Pin the Seed-TTS scorer to the second card: _get_eval_device() otherwise
+          # auto-picks npu:0, which the omni server under test already fills with its
+          # KV cache, so Whisper OOMs on `.to(device)` and every clip reports
+          # "ASR / WER failed".
+          - export SEED_TTS_WER_EVAL=1
+          - export SEED_TTS_EVAL_DEVICE=npu:1
           - |
             set +e
             pytest -s -v tests/e2e/accuracy/minicpmo_4_5/test_minicpmo_4_5.py -m "full_model and npu and A3" --run-level full_model
             EXIT=$$?
             buildkite-agent artifact upload "tests/e2e/accuracy/minicpmo_4_5/results/*.json"
             exit $$EXIT
+      - label: ":full_moon: Omni · MiniCPM-o 4.5 · Perf Test"
+        key: nightly-omni-performance-minicpmo-4-5-npu
+        timeout_in_minutes: 180
+        mirror_hardwares: a3_npu_2
+        env:
+          VLLM_WORKER_MULTIPROC_METHOD: spawn
+          VLLM_ALLOW_LONG_MAX_MODEL_LEN: "1"
+          HF_TOKEN: "${HF_TOKEN}"
+        commands:
+          - export BENCHMARK_DIR=tests/dfx/perf/results
+          - |
+            set +e
+            pytest -s -v tests/dfx/perf/scripts/run_benchmark.py -m "npu and A3" --test-config-file tests/dfx/perf/tests/test_minicpmo_4_5.json
+            EXIT=$$?
+            buildkite-agent artifact upload "tests/dfx/perf/results/*.json"
+            exit $$EXIT
+      - label: ":full_moon: Omni · MiniCPM-o 4.5 · Perf Test (A2)"
+        key: nightly-omni-performance-minicpmo-4-5-npu-a2
+        timeout_in_minutes: 180
+        mirror_hardwares: a2b3_npu_1
+        env:
+          VLLM_WORKER_MULTIPROC_METHOD: spawn
+          VLLM_ALLOW_LONG_MAX_MODEL_LEN: "1"
+          HF_TOKEN: "${HF_TOKEN}"
+          HF_HUB_OFFLINE: "1"
+        commands:
+          - export BENCHMARK_DIR=tests/dfx/perf/results
+          - |
+            set +e
+            pytest -s -v tests/dfx/perf/scripts/run_benchmark.py -m "npu and A2" --test-config-file tests/dfx/perf/tests/test_minicpmo_4_5.json
+            EXIT=$$?
+            buildkite-agent artifact upload "tests/dfx/perf/results/*.json"
+            exit $$EXIT
+      - label: ":full_moon: Omni · MiniCPM-o 4.5 · Duplex Seed-TTS Perf Test"
+        key: nightly-omni-performance-minicpmo-4-5-duplex-seed-tts-npu
+        timeout_in_minutes: 180
+        mirror_hardwares: a3_npu_2
+        env:
+          VLLM_WORKER_MULTIPROC_METHOD: spawn
+          VLLM_ALLOW_LONG_MAX_MODEL_LEN: "1"
+          HF_TOKEN: "${HF_TOKEN}"
+        commands:
+          - export BENCHMARK_DIR=tests/dfx/perf/results
+          - |
+            set +e
+            pytest -s -v tests/dfx/perf/scripts/run_benchmark.py -m "npu and A3" --test-config-file tests/dfx/perf/tests/test_minicpmo_4_5_duplex_seed_tts.json
+            EXIT=$$?
+            buildkite-agent artifact upload "tests/dfx/perf/results/*.json"
+            exit $$EXIT
   - group: ":card_index_dividers: TTS Model Test"
     key: nightly-tts-test-group
     depends_on: upload-nightly-pipeline
@@ -96,33 +151,6 @@ steps:
             EXIT=$$?
             buildkite-agent artifact upload "tests/dfx/perf/results/*.json"
             exit $$EXIT
-  - group: ":card_index_dividers: Diffusion X2I(&A&T) Model Test"
-    key: nightly-diffusion-x2iat-group
-    depends_on: upload-nightly-pipeline
-    if: >-
-      build.env("NIGHTLY") == "1" ||
-      build.pull_request.labels includes "nightly-test" ||
-      build.pull_request.labels includes "diffusion-x2iat-test"
-    steps:
-      - label: ":full_moon: Diffusion X2I(&A&T) · Perf Test · Qwen-Image-Edit-2511"
-        key: nightly-diffusion-x2iat-performance-qwen-image-edit-2511
-        timeout_in_minutes: 180
-        mirror_hardwares: a3_npu_8
-        env:
-          VLLM_WORKER_MULTIPROC_METHOD: spawn
-          VLLM_ALLOW_LONG_MAX_MODEL_LEN: "1"
-          HF_TOKEN: "${HF_TOKEN}"
-          DIFFUSION_ATTENTION_BACKEND: TORCH_SDPA
-        commands:
-          - export DIFFUSION_BENCHMARK_DIR=tests/dfx/perf/results
-          - export CACHE_DIT_VERSION=1.5.0
-          - |
-            set +e
-            pytest -s -v tests/dfx/perf/scripts/run_diffusion_benchmark.py -m npu --test-config-file tests/dfx/perf/tests/test_qwen_image_edit_2511_vllm_omni.json
-            EXIT=$$?
-            buildkite-agent artifact upload "tests/dfx/perf/results/diffusion_result_*.json"
-            buildkite-agent artifact upload "tests/dfx/perf/results/logs/*.log"
-            exit $$EXIT
   - group: ":card_index_dividers: Diffusion I2I Model Test"
     key: nightly-HunyuanImage3-x2i-test-group
     depends_on: upload-nightly-pipeline

.buildkite/npu/test-npu-ready.yml

@@ -1,4 +1,15 @@
 steps:
+  - label: "Platforms Test"
+    key: ready-platforms-test
+    depends_on: upload-ready-pipeline
+    mirror_hardwares: a3_npu_2
+    commands:
+      - >-
+        pytest -s -v
+        tests/platforms/npu/test_minicpmo_code2wav_npugraph.py
+        -m 'core_model and npu and A3 and omni'
+        --run-level 'core_model'
+
   - group: ":card_index_dividers: Omni Model Test"
     key: ready-omni-group
     depends_on: upload-ready-pipeline
@@ -32,7 +43,7 @@ steps:
     depends_on: upload-ready-pipeline
     steps:
       - label: ":sound: TTS · Qwen3-TTS CustomVoice Test"
-        mirror_hardwares: a2b3_npu_1
+        mirror_hardwares: a3_npu_2
         env:
           VLLM_WORKER_MULTIPROC_METHOD: spawn
           VLLM_ALLOW_LONG_MAX_MODEL_LEN: "1"

.claude/skills/add-diffusion-model/SKILL.md

@@ -610,6 +610,24 @@ See the [Profiling Single-Stage Diffusion](../../../docs/contributing/profiling.
 
 ---
 
+## Pre-commit conventions
+
+New library files must pass the local gates in
+[docs/contributing/README.md](../../../docs/contributing/README.md#linting).
+That page is the full hook list (SPDX, forbidden imports including Hugging Face
+Hub / Triton / pickle, `torch.cuda`, mypy, test marks, markdownlint, Buildkite,
+shellcheck). In particular:
+
+- SPDX copyright is `vLLM-Omni project` (stale `vLLM project` is rewritten).
+- Use `import regex as re` and `pybase64` in `vllm_omni/`; do not import stdlib
+  `re` or `base64`. Hugging Face Hub downloads go through
+  `vllm.transformers_utils.repo_utils`.
+- Do not add `torch.cuda.*` call sites; use `current_omni_platform`.
+- New `tests/**/test_*.py` files need a CI level mark and a hardware mark.
+- Do not expand `CHECK_IMPORTS[*].allowed_files` or `ALLOWED_FILES` without review.
+- GitHub Actions skips SPDX/shellcheck/mypy-3.10/test-marks/markdownlint; run
+  `pre-commit` locally.
+
 ## Iterative Development Tips
 
 1. **Start minimal**: Basic generation first, no parallelism/caching

.claude/skills/add-tts-model/SKILL.md

@@ -138,9 +138,12 @@ A shared buffer silently corrupts audio across concurrent requests — the sympt
    - Load codec weights (may need lazy loading from separate checkpoint)
    - Implement `forward()`: codec codes -> audio waveform
    - Return `OmniOutput` with `multimodal_outputs`
-5. **Create stage config YAML** defining both stages, memory allocation, and model paths
-6. **Create stage input processor** for prompt building
-7. **Write end2end.py** test script
+5. **Define and register the pipeline topology** in
+   `vllm_omni/model_executor/models/<model>/pipeline.py`
+6. **Create the deploy YAML** under `vllm_omni/deploy/` for placement,
+   memory sizing, connectors, and runtime overrides
+7. **Create stage input processor** for prompt building and inter-stage handoff
+8. **Write end2end.py** test script
 
 ### Critical Parameters to Get Right
 
@@ -174,7 +177,7 @@ These bugs appear in almost every new TTS PR. Check all before the first push. S
 - **All return paths must emit `model_outputs`** — if any early-return branch skips setting `model_outputs`, the serving layer silently drops that step's audio (fish speech: `fix: ensure ALL return paths emit model_outputs`)
 - **Per-request state isolation** — for batched concurrent requests, key all state by request ID; a shared buffer corrupts audio across requests (fish speech: `fix: per-request vocode + delta emission`)
 - **Codec tensor device** — move codec codes to the codec decoder's device before calling decode; mismatches cause silent CPU fallback or crashes (fish speech: `fix: use model device for CUDA stream`)
-- **AR stage `max_num_seqs`** — set to at least 4 in production configs; for single-stage models this is the only stage. For two-stage models, Stage 0 (AR) needs `max_num_seqs ≥ 4` to pipeline concurrent requests; Stage 1 (codec decoder) typically uses `max_num_seqs: 1` intentionally. Default of 1 everywhere causes audio gaps under concurrency because the codec window round-robins across requests (RFC #2568)
+- **AR stage `max_num_seqs`** — set to at least 4 in production deploy configs; for single-stage models this is the only stage. For two-stage models, Stage 0 (AR) needs `max_num_seqs ≥ 4` to pipeline concurrent requests; Stage 1 (codec decoder) is model-specific and may intentionally use `max_num_seqs: 1`. Defaulting the AR stage to 1 causes audio gaps under concurrency because the codec window round-robins across requests (RFC #2568)
 
 ### Optional Dependency Handling
 
@@ -190,18 +193,18 @@ pattern, signature constraints, and MOSS-TTS-Nano reference.
 When the upstream model cannot be cleanly split into an AR stage and a
 separate decoder, run the full pipeline inside a single AR worker and
 stream audio through a per-request `inference_stream()` generator keyed by
-`_omni_req_id`. Stage config must set `worker_type: ar`,
-`engine_output_type: audio`, `final_output: true`, `is_comprehension: true`,
-and `async_chunk: false` at the top level. Only extract params from
+`_omni_req_id`. Define one `StagePipelineConfig` with
+`execution_type=StageExecutionType.LLM_AR`, `engine_output_type="audio"`,
+`final_output=True`, and `owns_tokenizer=True`. Set `async_chunk: false` in
+the deploy YAML. Only extract params from
 `additional_information` that you actually forward, or pre-commit fails
 `ruff F841`.
 
 Full walkthrough with the complete `forward()` / `_create_stream_gen()`
-skeleton and stage-config fields:
+skeleton plus pipeline/deploy definitions:
 [references/single-stage-ar.md](references/single-stage-ar.md). For an
 in-tree reference, look for any single-stage AR model under
-`vllm_omni/model_executor/models/` — e.g. the MOSS-TTS-Nano integration when
-it lands.
+`vllm_omni/model_executor/models/`, such as MOSS-TTS-Nano.
 
 **VoxCPM2 is a different pattern** and should not reuse this skeleton — it
 runs the base LM under vLLM PagedAttention with external side-computation.
@@ -210,7 +213,8 @@ See `plan/voxcpm2_native_ar_design.md`.
 ### Deliverables
 
 - Model files in `vllm_omni/model_executor/models/<model_name>/`
-- Stage config YAML
+- Registered `pipeline.py` topology
+- Deploy YAML under `vllm_omni/deploy/`
 - Working `end2end.py` at `examples/offline_inference/text_to_speech/<model>/end2end.py`
 - New section in `examples/offline_inference/text_to_speech/README.md` (table row + per-model section). Do **not** create a top-level `examples/offline_inference/<model>/` dir or a per-model `README.md` inside `text_to_speech/<model>/` — the hub README is the documented surface and the mkdocs `generate_examples` hook only descends one level into `examples/<category>/`.
 
@@ -341,8 +345,8 @@ Heavier scenarios in the same file use **`advanced_model` only** (streaming, ext
 @pytest.mark.tts
 @hardware_test(res={"cuda": "L4"}, num_cards=1)
 @pytest.mark.parametrize("omni_server", tts_server_params, indirect=True)
-def test_voice_clone_en_non_streaming_001(omni_server, openai_client) -> None:
-    openai_client.send_audio_speech_request({...})
+def test_voice_clone_en_non_streaming_001(omni_server, online_client) -> None:
+    online_client.send_audio_speech_request({...})
 ```
 
 **L4 consolidation:** Prefer parametrized `OmniServerParams` rows (`default`, `async_chunk`, feature flags) in one expansion module rather than many merge-only files ([#1832](https://github.com/vllm-project/vllm-omni/issues/1832)).
@@ -359,8 +363,8 @@ def test_voice_clone_en_non_streaming_001(omni_server, openai_client) -> None:
 
 | Path | Fixture | Call |
 |------|---------|------|
-| Online `/v1/*` | `openai_client` | `openai_client.send_*_request(request_config)` |
-| Offline inference | `omni_runner_handler` | `omni_runner_handler.send_*_request(request_config)` |
+| Online `/v1/*` | `online_client` | `online_client.send_*_request(request_config)` |
+| Offline inference | `offline_client` | `offline_client.send_*_request(request_config)` |
 
 1. **Grep `runtime.py` first** — reuse `send_omni_request`, `send_diffusion_request`, `send_audio_speech_request` (online + offline Qwen-style TTS), `send_single_stage_tts_request` (Coqui XTTS / MOSS-TTS-Nano offline), etc.
 2. **No matching helper** → add `send_<feature>_request` (or `send_<route>_http_request` for negative/dfx) in **`runtime.py`** with general `assert_*` bundled inside, **then** call it from the test.
@@ -392,7 +396,7 @@ See [vllm-omni-test skill](../vllm-omni-test/SKILL.md) § **Runtime send helpers
   `/health` return non-200 until you're actually ready.
 - **Mark tests per the CI Levels table** — baseline smoke: `core_model` + `advanced_model`; heavier cases: `advanced_model` only; L4 expansion: `full_model`
 - **No per-model helper modules** — do not create `tests/helpers/{slug}.py`; keep constants and `request_config` payloads in the test file
-- **Online and offline e2e go through `runtime.py`** — `openai_client.send_audio_speech_request` (online); `omni_runner_handler.send_audio_speech_request` (Qwen-style offline) or `send_single_stage_tts_request` (single-stage offline). Add a new `send_*_request` in `runtime.py` when none fits; do not embed `omni.generate` or HTTP in tests
+- **Online and offline e2e go through `runtime.py`** — `online_client.send_audio_speech_request` (online); `offline_client.send_audio_speech_request` (Qwen-style offline) or `send_single_stage_tts_request` (single-stage offline). Add a new `send_*_request` in `runtime.py` when none fits; do not embed `omni.generate` or HTTP in tests
 
 ## Phase 4: Async Chunk (Streaming)
 
@@ -400,21 +404,29 @@ See [vllm-omni-test skill](../vllm-omni-test/SKILL.md) § **Runtime send helpers
 
 ### Steps
 
-1. **Update stage config YAML**:
+1. **Update the pipeline topology** so the producing stage declares its async
+   handoff processor with `async_chunk_process_next_stage_input_func`.
+2. **Update the deploy YAML** to enable async chunk and configure the connector:
    ```yaml
    async_chunk: true
-   codec_chunk_frames: 25      # frames per chunk
-   codec_left_context_frames: 25  # overlap for smooth boundaries
+
+   connectors:
+     connector_of_shared_memory:
+       name: SharedMemoryConnector
+       extra:
+         codec_streaming: true
+         codec_chunk_frames: 25
+         codec_left_context_frames: 25
    ```
-2. **Implement chunk handling in Stage 1**:
+3. **Implement chunk handling in Stage 1**:
    - Accept partial input (chunk of codec codes)
    - Handle left context for smooth audio boundaries
    - Return partial audio in `OmniOutput`
-3. **Test streaming**:
+4. **Test streaming**:
    - Verify audio quality matches non-streaming output
    - Check for artifacts at chunk boundaries
    - Measure TTFA (time to first audio)
-4. **Update online serving** to support `stream=true` with PCM output
+5. **Update online serving** to support `stream=true` with PCM output
 
 ### Streaming Architecture
 
@@ -435,7 +447,8 @@ Stage 0 (AR)                    Stage 1 (Decoder)
 
 ### Deliverables
 
-- Updated stage config with async_chunk enabled
+- Updated pipeline async handoff processor and deploy config with
+  `async_chunk: true`
 - Smooth streaming audio without boundary artifacts
 - TTFA metrics
 
@@ -507,14 +520,15 @@ Use this checklist when integrating a new TTS model:
 - [ ] Config classes created with `model_type` registration
 - [ ] Stage 0 (AR) implemented and generates correct tokens
 - [ ] Stage 1 (Decoder) produces correct audio from tokens — dtype float32 for codec decoder
-- [ ] Stage 1 `max_num_seqs` ≥ 4 in production config (default 1 causes gaps under concurrency)
+- [ ] AR stage `max_num_seqs` ≥ 4 in the production deploy config unless the model has a tested lower limit
 - [ ] Optional dependency fallbacks handled at `load_weights()` time (torchaudio/soundfile/etc.)
 - [ ] Streaming: codec codes accumulated across AR steps (not reset per step)
 - [ ] Streaming: delta audio emitted per chunk, not full re-decoded waveform
 - [ ] Streaming: all `forward()` return paths emit `model_outputs`
 - [ ] Streaming: per-request state keyed by request ID (not shared across requests)
 - [ ] Streaming: codec tensors moved to codec decoder device before decode
-- [ ] Stage config YAML created
+- [ ] Pipeline topology defined in `pipeline.py` and registered in `OMNI_PIPELINES`
+- [ ] Deploy YAML created under `vllm_omni/deploy/`
 - [ ] `end2end.py` produces audio matching reference quality
 - [ ] README.md written
 
@@ -531,7 +545,8 @@ Use this checklist when integrating a new TTS model:
 - [ ] Documentation added (offline + online docs, nav, supported models)
 
 ### Phase 4: Async Chunk
-- [ ] Stage config updated with `async_chunk: true`
+- [ ] Pipeline declares the async handoff processor
+- [ ] Deploy config sets `async_chunk: true` and connector chunk parameters
 - [ ] Stage 1 handles partial chunks correctly
 - [ ] No audio artifacts at chunk boundaries
 - [ ] Streaming via API (`stream=true`) works
@@ -548,7 +563,7 @@ Use this checklist when integrating a new TTS model:
 - [ ] `pre-commit run --files <changed>` passes before every push
 - [ ] Every commit has `Signed-off-by` matching the author email (`git commit -s`)
 - [ ] `git config user.email` matches the email registered on your GitHub account
-- [ ] Details and failure-recovery commands: [references/precommit-dco.md](references/precommit-dco.md)
+- [ ] Hook list, failure table, and DCO recovery: [references/precommit-dco.md](references/precommit-dco.md)
 
 ## References
 

.claude/skills/add-tts-model/references/precommit-dco.md

@@ -30,6 +30,18 @@ but the changes are correct — stage the modified files and re-commit.
 | `ruff F841` | Variable extracted but never forwarded to model call | Remove the extraction or wire it through |
 | `ruff E402` | Import added below function definitions | Move to top-level import block |
 | `ruff format` | Line length, spacing, quote style | Accept auto-fix, stage, re-commit |
+| `check-spdx-header` | Missing header, or copyright still says `vLLM project` | Two-line Omni header (`vLLM-Omni project`); restage the rewrite |
+| `check-forbidden-imports` | Stdlib `re`/`base64`, pickle, Hugging Face Hub API, or direct Triton/TileLang | `import regex as re` and `pybase64`; use `vllm.transformers_utils.repo_utils`. Do **not** add the file to `allowed_files` without review |
+| `check-torch-cuda-call` | New `torch.cuda.*` call site | Use `current_omni_platform` / `OmniPlatform`; do not grow `ALLOWED_FILES` without review |
+| `check-tts-adapter-migration` | New `self._tts_model_type` branch in `serving_speech.py` | Put per-model logic in `tts_adapters/`. Lower `MAX_MODEL_TYPE_BRANCHES` when removing branches; do not raise it without review |
+| `check-test-ci-coverage` | New `tests/**/test_*.py` missing level or hardware mark | Add `core_model`/`advanced_model`/… plus `cpu`/`cuda`/`hardware_test(` |
+| `mypy-3.10` | Type error on changed `vllm_omni/` files | Fix the types; model trees are excluded. Extra versions: `pre-commit run --hook-stage manual mypy-3.12` |
+| `markdownlint-cli2` | Docs/README markdown lint | Accept auto-fix under `docs/` / `recipes/` / root README |
+| `check-buildkite` | Invalid `.buildkite/*.yml` | Fix the YAML; do not grow `SKIP_FILES` without review |
+| `shellcheck` | No `shellcheck` on PATH, or script warning | `apt-get`/`dnf`/`brew install shellcheck`, or `shellcheck.exe` on Windows PATH |
+
+Canonical policy, CI `SKIP` list, and allowlist locations:
+[docs/contributing/README.md](../../../docs/contributing/README.md#linting).
 
 ## DCO sign-off
 

.claude/skills/add-tts-model/references/single-stage-ar.md

@@ -86,12 +86,51 @@ class YourModelForCausalLM(nn.Module):
 - Handle both `"audio"` (incremental) and `"result"` (final combined) event
   types from upstream models.
 
-## Stage config
+## Pipeline and deploy config
 
-Single stage with `worker_type: ar`, `engine_output_type: audio`,
-`final_output: true`, `is_comprehension: true`, and `async_chunk: false` at
-the top level. Omitting any of these causes silent misclassification in the
-serving layer.
+Define the fixed single-stage topology in `pipeline.py`:
+
+```python
+from vllm_omni.config.stage_config import (
+    PipelineConfig,
+    StageExecutionType,
+    StagePipelineConfig,
+)
+
+YOUR_TTS_PIPELINE = PipelineConfig(
+    model_type="your_model_name",
+    default_deploy_config_name="your_model_name.yaml",
+    model_arch="YourModelForCausalLM",
+    stages=(
+        StagePipelineConfig(
+            stage_id=0,
+            model_stage="your_model_name",
+            execution_type=StageExecutionType.LLM_AR,
+            input_sources=(),
+            owns_tokenizer=True,
+            engine_output_type="audio",
+            final_output=True,
+            final_output_type="audio",
+        ),
+    ),
+)
+```
+
+Register it in `vllm_omni/config/pipeline_registry.py`. Keep placement and
+runtime sizing in `vllm_omni/deploy/your_model_name.yaml`:
+
+```yaml
+async_chunk: false
+
+stages:
+  - stage_id: 0
+    devices: "0"
+    max_num_seqs: 4
+    gpu_memory_utilization: 0.8
+```
+
+Do not move `execution_type`, output ownership, tokenizer ownership, or model
+architecture into the deploy YAML; those are pipeline topology.
 
 ## Lint discipline
 
@@ -101,8 +140,5 @@ pre-commit.
 
 ## Reference implementation
 
-Look for any single-stage AR model under
-`vllm_omni/model_executor/models/` — e.g. `moss_tts_nano/` when its
-integration lands. If none is in tree yet, follow the skeleton above and
-cross-check against the `is_comprehension: true` / `async_chunk: false`
-dispatch in `vllm_omni/entrypoints/openai/serving_speech.py`.
+Use `vllm_omni/model_executor/models/moss_tts_nano/pipeline.py` and
+`vllm_omni/deploy/moss_tts_nano.yaml` as the in-tree reference.

.claude/skills/find-simplifications/SKILL.md

@@ -0,0 +1,135 @@
+---
+name: find-simplifications
+description: Find evidence-backed simplification candidates in vLLM-Omni and, when requested, turn them into focused proposals or code changes. Use for audits of dead, duplicated, speculative, over-generalized, unnecessarily defensive, or hand-rolled code. Use review-pr for ordinary correctness review and diffusion-perf-opt for performance-first optimization.
+---
+
+# Find vLLM-Omni Simplifications
+
+Find a small number of well-proven ways to reduce code, state, APIs, or maintenance cost without weakening supported behavior. Treat this as an architecture and ownership audit, not a dead-code search or style cleanup.
+
+The default deliverable is a read-only report. Do not add TODOs, design documents, issues, or code unless the user asks for those changes.
+
+## Establish the Contract
+
+Before judging a candidate:
+
+1. Read the repository instructions supplied by the host.
+2. Read [`docs/design/architecture_overview.md`](../../../docs/design/architecture_overview.md).
+3. Select only the relevant active module and feature documents from [`docs/design/index.md`](../../../docs/design/index.md). Archived design pages are historical context, not current contracts.
+4. Identify the comparison base with `git merge-base HEAD origin/main`; do not assume the current branch represents the intended baseline.
+5. If tests or CI coverage may change, use [`vllm-omni-test`](../vllm-omni-test/SKILL.md) for test placement, markers, and commands.
+
+Separate an implementation accident from an intentional boundary. In particular:
+
+- vLLM-Omni extends and sometimes overrides upstream vLLM. A similar upstream implementation is evidence for reuse only after comparing API, lifecycle, and multi-stage semantics.
+- Registries, import strings, model configuration, deployment YAML, plugins, and optional dependencies create dynamic consumers that a static reference count can miss.
+- Hardware backends, model families, execution modes, and connector implementations may be intentionally parallel. Do not propose collapsing one only because it is inactive in the local environment.
+- Stage, replica, request, cache, and transport state can look redundant while representing different owners. Prove that two fields or barriers encode the same fact before merging them.
+- A shorter hot path is not a simplification if it adds synchronization, tensor copies, host materialization, recompilation, or a measurable quality/performance regression.
+
+## Strong Candidates
+
+A strong candidate removes a meaningful surface and has a clear replacement or deletion boundary. Examples include:
+
+- A public method, configuration option, registry entry, route, message field, or event has no production consumer and no compatibility obligation.
+- Tests or documentation are the only consumers of behavior that is not part of a supported contract.
+- Multiple layers mirror the same request, lifecycle, cache-validity, readiness, or topology state.
+- AR, diffusion, platform, or model-specific implementations duplicate a stable common operation and can share it without hiding performance-critical differences.
+- An inherited upstream vLLM primitive is reimplemented locally without an Omni-specific semantic difference.
+- Compatibility or fallback branches protect environments the project no longer supports.
+- Serialization, IPC, output formatting, or connector code materializes the same payload more than once or converts between equivalent representations.
+- A helper, abstraction, package split, or extension point has only one real implementation and adds more indirection than policy.
+- Hand-written infrastructure can be replaced by an existing project dependency, Python standard-library feature, PyTorch primitive, or upstream vLLM facility with net deletion and equal observability.
+- Tests maintain large mocks, snapshots, or expected data solely for an unused API or representation.
+
+Do not elevate typo fixes, formatting, isolated renames, or vague complexity complaints into simplification candidates. Bundle small cleanups only when they support one substantive removal.
+
+## Survey by Ownership Boundary
+
+For broad audits, cover the largest or riskiest production deltas first. Useful domains are:
+
+- `vllm_omni/engine/`: orchestration, stage pools, request state, output convergence, cancellation, and replica lifecycle.
+- `vllm_omni/worker/` and AR model execution: upstream vLLM overlays, scheduler/cache state, collective RPC, adapters, and weight loading.
+- `vllm_omni/diffusion/`: request/step execution, batching, output materialization, IPC, offload, parallelism, and model pipelines.
+- `vllm_omni/entrypoints/`: duplicated validation/defaulting, offline/online parity, streaming, and route compatibility.
+- Connectors and distributed code: ownership, copies, barriers, handles, cleanup, and failure propagation across process/device/node boundaries.
+- Platform backends: genuine vendor differences versus copied control flow that has drifted.
+- Configuration, registries, recipes, and examples: duplicate sources of truth and options that never reach a runtime consumer.
+- Tests and CI: redundant fixtures, mocks of obsolete contracts, and expensive coverage that can move to a cheaper level without losing the guarded behavior.
+
+When the user explicitly requests a repository-wide or many-candidate audit and parallel agent work is available and authorized, divide these domains among agents and require the same evidence format from each. Otherwise survey them sequentially.
+
+## Prove or Reject Each Candidate
+
+Use `rg` first, then read every relevant caller and implementation. Search exact Python symbols plus serialized names, route paths, configuration keys, registry strings, YAML values, and CLI spellings.
+
+Classify evidence:
+
+- **Production:** `vllm_omni/`, runtime configuration, registries, serving entrypoints, connectors, recipes used for supported deployments, and executable loader paths.
+- **Validation/support:** `tests/`, benchmarks, examples, docs, CI, and developer tooling.
+- **External/dynamic:** upstream vLLM contracts, third-party plugins, model repositories, optional backends, and string-loaded objects. State what was checked and what remains unknown.
+
+For each candidate, answer:
+
+1. What exact surface would disappear or merge?
+2. Who owns the state or behavior today?
+3. Which production consumers exist, including dynamic consumers?
+4. What supported behavior, compatibility, performance, memory, or observability could change?
+5. What is the smallest validation that would catch a bad simplification?
+6. Is the result net deletion after tests, adapters, migration glue, and documentation are counted?
+
+Reject or downgrade the candidate when a live consumer exists and removal is actually a feature decision; when an active design contract explains the separation; when the change only moves complexity; or when hardware/model evidence is unavailable for a performance-sensitive claim.
+
+## Lifecycle and Data-Movement Audits
+
+For asynchronous or distributed code, map the ownership graph before recommending deletion:
+
+```text
+request admission -> orchestrator -> stage/replica -> worker/device
+  -> connector or IPC boundary -> output aggregation -> terminal cleanup
+```
+
+Associate every lock, event, sentinel, queue, readiness flag, timeout, abort path, and cache marker with an owner and transition. Coalesce mechanisms only when they represent the same state and have the same failure boundary. Preserve separate mechanisms when they protect publication versus rollback, local versus remote ownership, first-terminal-outcome arbitration, or cleanup after partial failure.
+
+For tensors and media, also trace device, process, and ownership transitions. Count GPU-to-CPU copies, shared-memory/object-store handles, serialization, decode/encode steps, and unlink/release responsibility. Prefer eliminating a transfer or representation over merely hiding it behind a helper.
+
+## Dependency Replacements
+
+Prefer facilities already available in the repository before adding a dependency. For a proposed replacement:
+
+- Name the exact local behavior it covers and any semantics that remain in glue code.
+- Check maintenance, license, optional-dependency boundaries, platform support, and transitive footprint.
+- Compare net deletion, test burden, import/startup cost, and runtime behavior.
+- Benchmark when the replaced code is on a model, scheduler, connector, serialization, or serving hot path.
+
+A wrapper that retains the same state machine or conversion logic is not a simplification.
+
+## Deliver the Result
+
+Report only candidates supported by concrete evidence. For each one include:
+
+- **Surface:** files and symbols involved.
+- **Evidence:** production consumers found, dynamic paths checked, and relevant design contract.
+- **Simplification:** exactly what to remove, merge, inline, or reuse.
+- **Benefit:** deleted API/state/code, reduced data movement, or reduced maintenance/test burden.
+- **Risk and validation:** behavior that could change and the narrowest useful tests or measurements.
+- **Confidence:** high or medium, with the unresolved evidence named for medium-confidence items.
+
+Also list representative ideas that were rejected when that prevents repeated investigation. Prefer three high-confidence candidates over a long speculative inventory.
+
+If the user asks for a durable design proposal, place it under `docs/design/feature/` and register it in `docs/design/index.md`; use current design-document conventions and keep implementation details proportional to the decision. If the user asks for implementation, make the smallest coherent change, update affected tests/docs, and preserve public compatibility unless the user explicitly accepts a breaking change.
+
+## Folding Work from Another Branch
+
+Diff the source branch against its own merge base with `origin/main`, not against the current feature branch. Port only independently justified candidates. Consolidate overlaps into the candidate with stronger evidence, and do not preserve a candidate count by carrying weaker duplicates.
+
+## Validation
+
+Match validation to the outgoing diff:
+
+- Documentation or skill changes: run the repository skill validator when available, check links/paths, search for stale source-specific terminology, and run `git diff --check`.
+- Python changes: run the narrowest relevant pytest coverage through the repository environment, then lint/format the changed files.
+- Distributed, model, hardware, performance, or accuracy changes: run available CPU/static checks and state the exact GPU/NPU/model gap; never present simulated evidence as device validation.
+- Before submission, use [`precheck-pr`](../precheck-pr/SKILL.md) for the branch-level self-review.
+
+When summarizing, distinguish checks that passed from checks that were not runnable, and state the snapshot or branch against which the audit was performed.

.claude/skills/precheck-pr/SKILL.md

@@ -67,6 +67,10 @@ Ask: "Quick mode or full mode?" Then walk the checklist for the detected PR type
 
 **Also run the Examples-Policy check on every PR** using [references/examples-policy.md](references/examples-policy.md). Inspect only Python paths introduced by the diff. Treat a new model-, checkpoint-, vendor-, or family-specific Python example as blocking; do not report pre-existing example debt.
 
+**Also run the Simplification pass on every PR** using [find-simplifications](../find-simplifications/SKILL.md). Keep it diff-scoped: inspect added or modified public APIs, state, abstractions, fallback and compatibility paths, data movement, and only the adjacent ownership needed to prove a candidate. Do not expand the pass into a repository-wide audit. Zero candidates is a valid pass; report only evidence-backed opportunities. Mark a candidate as blocking only when it proves a correctness or merge-readiness problem, such as newly unreachable code. Otherwise report it as a warning.
+
+**Also confirm local pre-commit gates** from [docs/contributing/README.md](../../../docs/contributing/README.md#linting). GitHub Actions `SKIP` does not prove they passed. Full new-hook list: Omni SPDX (`vLLM-Omni project`); forbidden imports (stdlib `re`/`base64`, pickle, Hugging Face Hub API, Triton/TileLang); no new `torch.cuda` call sites; shellcheck (macOS/Windows must install the binary); mypy-3.10; test files have CI level + hardware marks; TTS `_tts_model_type` branches in `serving_speech.py` must not increase; markdownlint on `docs/`; Buildkite YAML schema. ✗ for growing `allowed_files` / `ALLOWED_FILES` or raising `MAX_MODEL_TYPE_BRANCHES` without review.
+
 ### Step 5: Print Report
 
 ```
@@ -80,6 +84,7 @@ Pre-check report for <branch>
   PR title format    ✓
   Code quality       ⚠ 1 broad except, 2 Any hints
   Examples policy    ✗ new model-specific Python example
+  Simplification     ⚠ helper duplicates an existing owner
   PR desc integrity  ✓
   Registry/config    ✓
   Dead code          ⚠ 2 warnings

.claude/skills/precheck-pr/references/checklists.md

@@ -164,6 +164,7 @@ Diffusion models live under `vllm_omni/diffusion/`. In addition to the New Model
 - [ ] **PR body matches diff:** description doesn't claim changes not in the diff
 - [ ] **Branch is rebased:** `git merge-base HEAD origin/main` is recent, no merge conflicts
 - [ ] **CI gates passing:** DCO, pre-commit, build — check `gh pr view --json statusCheckRollup` if a PR already exists
+- [ ] **Local pre-commit gates:** full list in [Linting](../../../../docs/contributing/README.md#linting). Blocking when the diff introduces: stale `vLLM project` SPDX; stdlib `re`/`base64` or pickle in library code; new `torch.cuda` call sites; `tests/**/test_*.py` without level+hardware marks; new `_tts_model_type` branches in `serving_speech.py`; invalid `.buildkite` YAML. macOS/Windows need a native `shellcheck`. ✗ for growing `CHECK_IMPORTS[*].allowed_files`, `ALLOWED_FILES`, or `MAX_MODEL_TYPE_BRANCHES` without review.
 
 ### Full (adds)
 

.claude/skills/readme.md

@@ -31,9 +31,12 @@ self-check.
 - [`diffusion-perf-opt`](diffusion-perf-opt/SKILL.md): guides diffusion model
   performance optimization, including profiling traces, parallel strategies,
   stage timing analysis, and benchmark-driven tuning
+- [`find-simplifications`](find-simplifications/SKILL.md): finds evidence-backed
+  opportunities to remove or merge dead, duplicated, speculative,
+  over-generalized, or unnecessarily defensive vLLM-Omni code
 - [`precheck-pr`](precheck-pr/SKILL.md): self-checks a branch before creating a
-  PR by validating title format, dead code, accuracy and performance claims,
-  and merge readiness
+  PR by validating title format, dead code, simplification opportunities,
+  accuracy and performance claims, and merge readiness
 - [`quantization`](quantization/SKILL.md): guides quantization method selection,
   model integration, checkpoint loading, and quality/performance validation
   for vLLM-Omni

.claude/skills/review-pr/SKILL.md

@@ -21,7 +21,10 @@ Make every finding:
 - **Calibrated:** match severity to user and maintainer impact.
 
 Do not report unrelated backlog, style already enforced by pre-commit, or a
-missing test that would not protect changed behavior.
+missing test that would not protect changed behavior. Do report new allowlist
+or budget entries in `check_forbidden_imports.py` / `check_torch_cuda.py` /
+`check_tts_adapter.py` / `check_buildkite.py` unless the PR justifies them:
+those are policy changes, not lint noise.
 
 ## Select the input and depth
 
@@ -92,12 +95,13 @@ If docs and live code disagree, verify the code/tests and report the drift.
 
 | Reference | Read when |
 | --- | --- |
-| [model-addition-checklist.md](references/checks/model-addition-checklist.md) | A model, architecture, loader, processor, registry, or stage config is added. |
+| [model-addition-checklist.md](references/checks/model-addition-checklist.md) | A model, architecture, loader, processor, registry, pipeline config, or deploy config is added. |
 | [perf-verification.md](references/checks/perf-verification.md) | The PR makes a latency, throughput, memory, or quality claim. |
 | [test-quality-evaluation.md](references/checks/test-quality-evaluation.md) | Tests change, are absent for risky code, or may not exercise production behavior. |
 | [tests-docs-checklist.md](references/checks/tests-docs-checklist.md) | Coverage, CI markers, examples, user docs, or PR evidence need review. |
 | [verification.md](references/checks/verification.md) | Hardware, a server, or a runnable affected path is available for active verification. |
 | [examples-policy.md](../precheck-pr/references/examples-policy.md) | The PR adds, copies, or renames Python under `examples/`; apply the canonical policy shared with `precheck-pr`. |
+| [find-simplifications](../find-simplifications/SKILL.md) | Every review; run a diff-scoped subtraction and simplification pass after correctness blockers. |
 
 ### Delivery and reviewer coordination
 
@@ -180,9 +184,13 @@ promotion gate. Candidate or draft rules are questions, not blockers, unless
 current code, tests, or policy enforce them. Inspect both sides of any config,
 registry, serialization, connector, cache, or stage boundary.
 
-When a diff adds or expands a helper, class, fallback, compatibility branch, or
-public behavior, run a subtraction pass: remove out-of-scope behavior and check
-whether each new abstraction can be deleted, merged, moved, or inlined.
+Read and apply [find-simplifications](../find-simplifications/SKILL.md) on every
+review. Constrain it to the diff and the adjacent ownership, callers, or
+consumers needed to prove a candidate. Check whether added or expanded helpers,
+classes, state, fallback and compatibility branches, data movement, or public
+behavior can be deleted, merged, moved, or inlined. Zero candidates is a valid
+result. Do not widen the review into repository backlog or report speculative
+style preferences as simplification findings.
 
 ### 6. Verify the changed path
 

.claude/skills/review-pr/references/checks/model-addition-checklist.md

@@ -1,7 +1,7 @@
 # Model Addition Checklist
 
 Read this reference when a PR adds or expands a model architecture, pipeline,
-loader, processor, registry entry, or stage configuration.
+loader, processor, registry entry, pipeline config, or deploy config.
 
 Official docs: [model contribution guides](https://docs.vllm.ai/projects/vllm-omni/en/latest/contributing/model/),
 [adding an omni model](https://docs.vllm.ai/projects/vllm-omni/en/latest/contributing/model/adding_omni_model/),

.claude/skills/review-pr/references/modules/configuration.md

@@ -3,7 +3,7 @@
 Primary design: [vLLM-Omni Configuration](https://docs.vllm.ai/projects/vllm-omni/en/latest/design/module/vllm_omni_config/).
 
 Use for config construction, deploy YAML, registries, schema, defaults,
-aliases, CLI projection, stage config, and topology. Inspect the reviewed
+aliases, CLI projection, pipeline config, and topology. Inspect the reviewed
 page's status: while its contract remains deferred or draft, current code and
 tests—not proposed precedence rules—are authoritative.
 

.claude/skills/review-pr/references/process/review-execution.md

@@ -111,6 +111,17 @@ Mark early findings as preliminary and continue. This is not a GitHub comment.
   but do not publish review events without separate authorization.
 - Record DCO, pre-commit, required CI, and mergeability. Pending or unknown gates
   do not block source review.
+- GitHub Actions `SKIP` omits SPDX, shellcheck, markdownlint, mypy-3.10, and
+  test-mark coverage. A green GHA pre-commit job does not prove those local
+  gates passed. New files still need the full Linting list: Omni SPDX
+  (`vLLM-Omni project`), no stdlib `re`/`base64` in `vllm_omni/`, no new
+  pickle / Hugging Face Hub API / `torch.cuda` call sites, test marks, TTS
+  adapter ratchet, Buildkite schema, and native shellcheck on macOS/Windows.
+  See [Linting](https://vllm-omni.readthedocs.io/en/latest/contributing/#linting).
+- Expanding `CHECK_IMPORTS[*].allowed_files`, `ALLOWED_FILES`,
+  `MAX_MODEL_TYPE_BRANCHES`, or Buildkite `SKIP_FILES` is a policy change.
+  Do not rubber-stamp allowlist/budget growth; require justification and
+  prefer fixing the call site.
 - Treat a failed gate as its own evidence. Do not restate its formatting/lint
   output as a new code-review finding.
 - Open CI logs only when the first failing step overlaps the frozen diff or

.claude/skills/review-pr/references/process/review-routing.md

@@ -48,7 +48,7 @@ compatibility changes.
 
 | Signal | Read | Optional repo-local skill |
 | --- | --- | --- |
-| New or expanded model, loader, processor, registry, or stage config | [model-addition-checklist.md](../checks/model-addition-checklist.md) | [`add-tts-model`](../../../add-tts-model/SKILL.md) or [`add-diffusion-model`](../../../add-diffusion-model/SKILL.md) |
+| New or expanded model, loader, processor, registry, pipeline config, or deploy config | [model-addition-checklist.md](../checks/model-addition-checklist.md) | [`add-tts-model`](../../../add-tts-model/SKILL.md) or [`add-diffusion-model`](../../../add-diffusion-model/SKILL.md) |
 | Latency, throughput, memory, scaling, precision, or quality claim | [perf-verification.md](../checks/perf-verification.md) | [`diffusion-perf-opt`](../../../diffusion-perf-opt/SKILL.md) for diffusion |
 | Tests changed, absent for risky behavior, or test-only | [test-quality-evaluation.md](../checks/test-quality-evaluation.md) | [`vllm-omni-test`](../../../vllm-omni-test/SKILL.md) |
 | CI, examples, docs, public behavior, or contributor evidence | [tests-docs-checklist.md](../checks/tests-docs-checklist.md) | None |

.claude/skills/vllm-omni-test/SKILL.md

@@ -94,8 +94,8 @@ Existing references: `tests/e2e/offline_inference/test_qwen2_5_omni.py` (L2-styl
 
 | Scenario | Typical location | Fixtures / runner pattern | Baseline markers & level |
 |----------|------------------|---------------------------|---------------------------|
-| **Offline inference e2e** | `tests/e2e/offline_inference/` | **Module (default):** `omni_runner` + `omni_runner_handler`. **Function (isolation only):** `omni_runner_function` + `omni_runner_handler_function`. Diffusion/TTS may use `Omni(...).generate` directly | L2: `core_model` + **one of** `omni` / `tts` / `diffusion`; `@hardware_test(...)` when GPU/NPU is required |
-| **Online serving e2e** | `tests/e2e/online_serving/` | **Module (default):** `omni_server` + `openai_client`. **Function (isolation only):** `omni_server_function` + `openai_client_function`. Clients: `send_omni_request` (omni), `send_audio_speech_request` (tts), `send_diffusion_request` / `send_video_diffusion_request` / `send_images_generations_request` (diffusion) | Baseline smoke: **`core_model` + `advanced_model`**; heavier paths: `advanced_model` only; L4 expansion: `full_model` |
+| **Offline inference e2e** | `tests/e2e/offline_inference/` | **Module (default):** `omni_runner` + `offline_client`. **Function (isolation only):** `omni_runner_function` + `offline_client_function`. Diffusion/TTS may use `Omni(...).generate` directly | L2: `core_model` + **one of** `omni` / `tts` / `diffusion`; `@hardware_test(...)` when GPU/NPU is required |
+| **Online serving e2e** | `tests/e2e/online_serving/` | **Module (default):** `omni_server` + `online_client`. **Function (isolation only):** `omni_server_function` + `online_client_function`. Clients: `send_omni_request` (omni), `send_audio_speech_request` (tts), `send_diffusion_request` / `send_video_diffusion_request` / `send_images_generations_request` (diffusion) | Baseline smoke: **`core_model` + `advanced_model`**; heavier paths: `advanced_model` only; L4 expansion: `full_model` |
 | **Documentation / runnable examples** | `tests/examples/offline_inference/`, `tests/examples/online_serving/` | **Offline docs (preferred):** extract Python/Bash blocks from the doc README (e.g. `ReadmeSnippet.extract_readme_snippets`), `pytest.mark.parametrize` each snippet, run via `example_runner.run` with a stable `output_subfolder`. **Online docs:** copy client/request scripts into dedicated tests and keep them in sync with the doc page. | Usually **L4**: `advanced_model`, often `example` plus hardware marks matching the nightly docs-example job (see `.buildkite/cuda/test-nightly.yml`). Full conventions: [docs/contributing/ci/test_examples/l4_doc_example_tests.inc.md](../../../docs/contributing/ci/test_examples/l4_doc_example_tests.inc.md) (introduced in [PR #1910](https://github.com/vllm-project/vllm-omni/pull/1910): naming, output directory layout, skip rules, avoid trimming `num_inference_steps` without a strong CI reason). |
 | **Performance / benchmark** | `tests/dfx/perf/tests/*.json` + `run_*_benchmark.py` | JSON or script-driven server + load config; assert explicit metrics / baselines | L4 Perf: JSON `mark` with `full_model` + `omni`/`tts`/`diffusion`; wire `test-nightly.yml` Perf steps |
 | **Invalid parameter / negative HTTP validation** | `tests/dfx/reliability/invalid_param_test/` | Live `omni_server` + low-level `send_*_http_request` with `err_code` / `err_message` | `pytest.mark.slow` + `omni` / `tts` / `diffusion` + `@hardware_marks` (`H100` or `L4`); CI in **`test-weekly.yml`** (not ready/merge/nightly) |
@@ -110,10 +110,10 @@ After choosing offline/online/docs/perf, classify the **product under test** and
 |-----------|------------------------------|----------------------------|----------------------------------------|
 | **What it is** | Multimodal LLM pipeline (thinker/talker/stages; text + vision + audio I/O) | Speech synthesis / voice models | Generative diffusion (noise → image, audio, text, or video) |
 | **Examples** | Qwen2.5-Omni, Qwen3-Omni | Qwen3-TTS, VoxCPM2, Higgs-Audio, Step-Audio2 | Qwen-Image, BAGEL, Wan2.2, HunyuanVideo |
-| **Offline runner** | `omni_runner` + `omni_runner_handler`, `generate_multimodal` | `Omni(...).generate` or stage YAML + TTS params | `Omni(...).generate` + `OmniDiffusionSamplingParams` |
-| **Online client** | `openai_client.send_omni_request` (chat completions, modalities) | `openai_client.send_audio_speech_request` (`/v1/audio/speech`) | `send_diffusion_request` (chat/T2I), `send_video_diffusion_request` (`/v1/videos`, X2V), or `send_images_generations_http_request` / `send_images_edits_http_request` (DALL-E routes) |
-| **Typical assertions** | Stage outputs, text/audio keywords via `OmniRunnerHandler` / response handler | WAV bytes, stream chunks, speech endpoint contract | Image/video dimensions, `final_output_type`, `assert_diffusion_response` |
-| **Stage / deploy YAML** | Per-model omni stage configs (`ci/qwen3_omni_moe.yaml`, …) | `qwen3_tts.yaml`, `voxcpm2.yaml`, … | Often default serve; parallel/offload YAML for heavy DiT |
+| **Offline runner** | `omni_runner` + `offline_client`, `generate_multimodal` | `Omni(...).generate` with deploy YAML + TTS params | `Omni(...).generate` + `OmniDiffusionSamplingParams` |
+| **Online client** | `online_client.send_omni_request` (chat completions, modalities) | `online_client.send_audio_speech_request` (`/v1/audio/speech`) | `send_diffusion_request` (chat/T2I), `send_video_diffusion_request` (`/v1/videos`, X2V), or `send_images_generations_http_request` / `send_images_edits_http_request` (DALL-E routes) |
+| **Typical assertions** | Stage outputs, text/audio keywords via `OfflineOmniClient` / response handler | WAV bytes, stream chunks, speech endpoint contract | Image/video dimensions, `final_output_type`, `assert_diffusion_response` |
+| **Deploy YAML** | Per-model deploy configs (`ci/qwen3_omni_moe.yaml`, …) | `qwen3_tts.yaml`, `voxcpm2.yaml`, … | Often default serve; parallel/offload YAML for heavy DiT |
 | **Nightly group** (`test-nightly.yml`) | **Omni Model Test** — `-m "full_model and omni"` | **TTS Model Test** — `-m "full_model and tts"` | **Diffusion X2I(&A&T)** *or* **Diffusion X2V** (see Step 3 table; same `diffusion` marker, different YAML group / file shard) |
 | **L4 pressure** | Expansion per modality/model as needed | Expansion + accuracy/perf in TTS group | X2I: merge feature combos per [#1832](https://github.com/vllm-project/vllm-omni/issues/1832); X2V: separate nightly group |
 
@@ -175,9 +175,9 @@ See **L1 unit test constraints (mocking)** below for the full do/don't list.
 @pytest.mark.omni
 @hardware_test(...)
 @pytest.mark.parametrize("omni_runner", test_params, indirect=True)
-def test_<scenario>(omni_runner, omni_runner_handler) -> None:
+def test_<scenario>(omni_runner, offline_client) -> None:
     request_config = {"prompts": ..., "modalities": [...]}  # optional: images, videos, audios
-    omni_runner_handler.send_omni_request(request_config)
+    offline_client.send_omni_request(request_config)
 ```
 
 *Offline generative e2e — **Diffusion** (representative):*
@@ -187,8 +187,8 @@ def test_<scenario>(omni_runner, omni_runner_handler) -> None:
 @pytest.mark.diffusion
 @hardware_test(...)
 @pytest.mark.parametrize("omni_runner", test_params, indirect=True)
-def test_text_to_image_001(omni_runner_handler) -> None:
-    omni_runner_handler.send_diffusion_request({"prompt": "...", "extra_body": {"num_inference_steps": 4, ...}})
+def test_text_to_image_001(offline_client) -> None:
+    offline_client.send_diffusion_request({"prompt": "...", "extra_body": {"num_inference_steps": 4, ...}})
 ```
 
 *Offline TTS e2e — **Qwen3-TTS** (two-stage; representative):*
@@ -198,8 +198,8 @@ def test_text_to_image_001(omni_runner_handler) -> None:
 @pytest.mark.tts
 @hardware_test(...)
 @pytest.mark.parametrize("omni_runner", tts_server_params, indirect=True)
-def test_text_to_audio_001(omni_runner, omni_runner_handler) -> None:
-    omni_runner_handler.send_audio_speech_request({
+def test_text_to_audio_001(omni_runner, offline_client) -> None:
+    offline_client.send_audio_speech_request({
         "input": "...",
         "task_type": "Base",
         "ref_audio": REF_AUDIO_URL,
@@ -214,8 +214,8 @@ def test_text_to_audio_001(omni_runner, omni_runner_handler) -> None:
 @pytest.mark.tts
 @hardware_test(...)
 @pytest.mark.parametrize("omni_runner", tts_server_params, indirect=True)
-def test_voice_clone_001(omni_runner_handler) -> None:
-    omni_runner_handler.send_single_stage_tts_request({
+def test_voice_clone_001(offline_client) -> None:
+    offline_client.send_single_stage_tts_request({
         "input": "...",
         "language": "en",
         "prompt_audio_path": REF_AUDIO_PATH,
@@ -231,9 +231,9 @@ def test_voice_clone_001(omni_runner_handler) -> None:
 @pytest.mark.omni
 @hardware_test(...)
 @pytest.mark.parametrize("omni_server", test_params, indirect=True)
-def test_text_to_text_001(omni_server, openai_client) -> None:
+def test_text_to_text_001(omni_server, online_client) -> None:
     request_config = {"model": omni_server.model, "messages": ..., "modalities": ["text"]}
-    openai_client.send_omni_request(request_config)
+    online_client.send_omni_request(request_config)
 ```
 
 *Online serving e2e — **TTS** (representative):*
@@ -244,9 +244,9 @@ def test_text_to_text_001(omni_server, openai_client) -> None:
 @pytest.mark.tts
 @hardware_test(...)
 @pytest.mark.parametrize("omni_server", tts_server_params, indirect=True)
-def test_text_to_audio_001(omni_server, openai_client) -> None:
+def test_text_to_audio_001(omni_server, online_client) -> None:
     request_config = {"model": omni_server.model, "input": "...", "response_format": "wav", ...}
-    openai_client.send_audio_speech_request(request_config)
+    online_client.send_audio_speech_request(request_config)
 ```
 
 *Online serving e2e — **Diffusion** X2I (representative):*
@@ -255,8 +255,8 @@ def test_text_to_audio_001(omni_server, openai_client) -> None:
 @pytest.mark.core_model
 @pytest.mark.diffusion
 @pytest.mark.parametrize("omni_server", _get_default_case(MODEL), indirect=True)
-def test_text_to_image_001(omni_server, openai_client) -> None:
-    openai_client.send_diffusion_request({...})  # chat completions + extra_body
+def test_text_to_image_001(omni_server, online_client) -> None:
+    online_client.send_diffusion_request({...})  # chat completions + extra_body
 ```
 
 *Online serving e2e — **Diffusion** X2V (representative):*
@@ -264,8 +264,8 @@ def test_text_to_image_001(omni_server, openai_client) -> None:
 ```python
 @pytest.mark.core_model
 @pytest.mark.diffusion
-def test_text_to_video_001(omni_server, openai_client) -> None:
-    openai_client.send_video_diffusion_request({"model": ..., "form_data": {...}})  # /v1/videos
+def test_text_to_video_001(omni_server, online_client) -> None:
+    online_client.send_video_diffusion_request({"model": ..., "form_data": {...}})  # /v1/videos
 ```
 
 *Documentation example tests:* follow the **Preferred Test Strategy** in [l4_doc_example_tests.inc.md](../../../docs/contributing/ci/test_examples/l4_doc_example_tests.inc.md): dynamic extraction for offline READMEs; explicit copied client code for online pages until extraction is justified; use the documented **naming**, **output directory** (page folder + case id), and **skipping** rules (e.g. Gradio-only scripts).
@@ -274,6 +274,7 @@ def test_text_to_video_001(omni_server, openai_client) -> None:
 
 **3. Cross-cutting rules**
 
+- New `.py` files need the Omni SPDX header (`Copyright contributors to the vLLM-Omni project`, not `vLLM project`). `tests/` may keep stdlib `re`/`base64`; pickle is still banned unless the file is already on the pickle allowlist. Every collected `tests/**/test_*.py` needs a CI **level** mark and a **hardware** mark/helper (`check-test-ci-coverage` runs locally; GHA skips it). Do not grow `allowed_files` to land a test. Policy: [docs/contributing/README.md](../../../docs/contributing/README.md#linting).
 - Reuse existing fixtures for the chosen scenario; do not mix “online client” assumptions into offline `OmniRunner` tests without a clear reason.
 - Avoid external network dependency in assertions unless the scenario is explicitly “online serving” or doc examples that require a model hub (then align with CI secrets/cache).
 - Keep **one test function = one intent** (one modality combo, one endpoint contract, or one acceleration combo).
@@ -323,21 +324,21 @@ vLLM-Omni e2e tests start a real **OmniServer** (online) or **OmniRunner** (offl
 
 | Scope | Online fixtures | Offline fixtures | When to use |
 |-------|-----------------|------------------|-------------|
-| **Module** (default) | `omni_server` → `openai_client` | `omni_runner` → `omni_runner_handler` | **Default for L2/L3/L4 expansion** — amortize model/server init across `test_*` in the same module. Same `OmniServerParams` / runner config can be reused by multiple tests. |
-| **Function** | `omni_server_function` → `openai_client_function` | `omni_runner_function` → `omni_runner_handler_function` | **Only when required** — each `test_*` must get a **clean** server/runner (no shared engine/GPU state). Typical: `tests/dfx/reliability/`, sleep/wakeup, crash/restart, tests that mutate global server state. |
+| **Module** (default) | `omni_server` → `online_client` | `omni_runner` → `offline_client` | **Default for L2/L3/L4 expansion** — amortize model/server init across `test_*` in the same module. Same `OmniServerParams` / runner config can be reused by multiple tests. |
+| **Function** | `omni_server_function` → `online_client_function` | `omni_runner_function` → `offline_client_function` | **Only when required** — each `test_*` must get a **clean** server/runner (no shared engine/GPU state). Typical: `tests/dfx/reliability/`, sleep/wakeup, crash/restart, tests that mutate global server state. |
 
 **Rules:**
 
 1. **Default to module scope** (`omni_server` / `omni_runner`) unless the scenario **explicitly** needs a fresh instance per test function.
-2. **Indirect parametrize name must match the fixture name**: `@pytest.mark.parametrize("omni_server", ...)` with `omni_server` + `openai_client`; `@pytest.mark.parametrize("omni_server_function", ...)` with `omni_server_function` + `openai_client_function`. Do not mix module fixture with function client (or vice versa).
+2. **Indirect parametrize name must match the fixture name**: `@pytest.mark.parametrize("omni_server", ...)` with `omni_server` + `online_client`; `@pytest.mark.parametrize("omni_server_function", ...)` with `omni_server_function` + `online_client_function`. Do not mix module fixture with function client (or vice versa).
 3. Different `OmniServerParams` per `test_*` (e.g. default vs `--enable-cpu-offload`) is still OK with **module** `omni_server` — pytest parametrizes per test node; only switch to `_function` when isolation between tests matters, not merely because `server_args` differ.
 4. One `test_*` with **many** server configs (expansion matrix) → single function + `@pytest.mark.parametrize("omni_server", [...], indirect=True)` + module `omni_server` (see in-tree `test_qwen_image_expansion.py`).
 
 ```python
 # Default — module-scoped server (L4 expansion)
 @pytest.mark.parametrize("omni_server", [pytest.param(OmniServerParams(model=MODEL), marks=H100)], indirect=True)
-def test_foo_images_generations_default_1024(omni_server, openai_client) -> None:
-    openai_client.send_images_generations_request({...})
+def test_foo_images_generations_default_1024(omni_server, online_client) -> None:
+    online_client.send_images_generations_request({...})
 
 # Function-scoped — reliability / per-test clean state only
 @pytest.mark.parametrize(
@@ -345,9 +346,9 @@ def test_foo_images_generations_default_1024(omni_server, openai_client) -> None
     [pytest.param(OmniServerParams(model=MODEL), marks=H100)],
     indirect=True,
 )
-def test_foo_sleep_wakeup_cycle(omni_server_function, openai_client_function) -> None:
-    openai_client_function.send_omni_sleep_http_request({...})
-    openai_client_function.send_omni_wakeup_http_request({...})
+def test_foo_sleep_wakeup_cycle(omni_server_function, online_client_function) -> None:
+    online_client_function.send_omni_sleep_http_request({...})
+    online_client_function.send_omni_wakeup_http_request({...})
 ```
 
 #### Runtime send helpers — online **and** offline (`tests/helpers/runtime.py`)
@@ -356,19 +357,19 @@ def test_foo_sleep_wakeup_cycle(omni_server_function, openai_client_function) ->
 
 | Principle | Action |
 |-----------|--------|
-| **Reuse first** | Grep `runtime.py` for an existing **`send_*_request`** on `OpenAIClientHandler` (online) or `OmniRunnerHandler` (offline) that matches the endpoint / pipeline shape. |
+| **Reuse first** | Grep `runtime.py` for an existing **`send_*_request`** on `OnlineOmniClient` (online) or `OfflineOmniClient` (offline) that matches the endpoint / pipeline shape. |
 | **Extend when close** | If an existing `send_*_request` almost fits (e.g. missing one optional field), extend it in `runtime.py` — do not fork logic in the test file. |
 | **Add when missing** | No suitable helper → add **`send_<feature>_request`** (high-level: call + general `assert_*`) or **`send_<route>_<verb>_http_request`** (low-level HTTP for negative/dfx) in **`runtime.py` first**, then call it from tests. |
 | **Test module owns payload only** | In `test_*.py`: `MODEL`, deploy path, vendored media, `get_prompt()`, and inline **`request_config` dicts** only. **No** `omni.generate(...)`, raw `requests.post`, OpenAI SDK calls, or `_collect_audio()` / `_process_output()` in e2e tests. |
 
-**Online** (`openai_client` from `omni_server`): `OpenAIClientHandler.send_*_request`.
+**Online** (`online_client` from `omni_server`): `OnlineOmniClient.send_*_request`.
 
-**Offline** (`omni_runner_handler` from `omni_runner`): `OmniRunnerHandler.send_*_request`.
+**Offline** (`offline_client` from `omni_runner`): `OfflineOmniClient.send_*_request`.
 
 | Do | Don't |
 |----|-------|
-| Online: `openai_client.send_omni_request`, `send_diffusion_request`, `send_audio_speech_request`, `send_video_diffusion_request`, `send_images_generations_request`, … | `requests.post(f"{base_url}/v1/...", json=…)` or `client.chat.completions.create(...)` inside a test |
-| Offline: `omni_runner_handler.send_omni_request`, `send_diffusion_request`, `send_audio_speech_request`, `send_single_stage_tts_request`, `send_single_stage_tts_batch_request`, … | `omni_runner.omni.generate(...)` + hand-rolled tensor/WAV extraction in `test_*.py` |
+| Online: `online_client.send_omni_request`, `send_diffusion_request`, `send_audio_speech_request`, `send_video_diffusion_request`, `send_images_generations_request`, … | `requests.post(f"{base_url}/v1/...", json=…)` or `client.chat.completions.create(...)` inside a test |
+| Offline: `offline_client.send_omni_request`, `send_diffusion_request`, `send_audio_speech_request`, `send_single_stage_tts_request`, `send_single_stage_tts_batch_request`, … | `omni_runner.omni.generate(...)` + hand-rolled tensor/WAV extraction in `test_*.py` |
 | Add missing `send_*` to **`runtime.py`** first; bundle general `assert_*` inside the send helper | A one-off `def _post_*` / `def _collect_audio` at the bottom of a test module |
 | Mirror naming/style of neighboring `send_*` (docstring, `request_config` dict, optional `run_level`, `err_code` / `err_message` for negative cases) | Different parameter shapes per test file for the same endpoint |
 
@@ -391,7 +392,7 @@ def test_foo_sleep_wakeup_cycle(omni_server_function, openai_client_function) ->
 
 Changing `request_config` fields (`size`, `n`, `seed`, server flags) is **not** special validation — the general `assert_*` should read those from `request_config` inside `send_*`.
 
-**`send_*` ↔ `assert_*` pairing (online `OpenAIClientHandler`):**
+**`send_*` ↔ `assert_*` pairing (online `OnlineOmniClient`):**
 
 | High-level `send_*` (prefer in L2+ e2e) | Assert already invoked inside `runtime.py` | Low-level HTTP-only sibling |
 |----------------------------------------|--------------------------------------------|-----------------------------|
@@ -404,7 +405,7 @@ Changing `request_config` fields (`size`, `n`, `seed`, server flags) is **not**
 
 **Rule:** Tests call high-level `send_*_request` for the general contract. **Never** call the same bundled `assert_*` again. Call an extra `assert_*` in the test **only** for special, case-specific validation.
 
-**Common `OpenAIClientHandler` entry points** (non-exhaustive — grep `runtime.py` before adding):
+**Common `OnlineOmniClient` entry points** (non-exhaustive — grep `runtime.py` before adding):
 
 | Area | High-level (SDK + assert) | Low-level HTTP (`*_http_request`) |
 |------|---------------------------|-----------------------------------|
@@ -415,7 +416,7 @@ Changing `request_config` fields (`size`, `n`, `seed`, server flags) is **not**
 | TTS (online) | `send_audio_speech_request` | `send_audio_speech_http_request`, `send_audio_generate_http_request`, … |
 | Ops / meta | — | `send_health_http_request`, `send_models_http_request`, `send_omni_sleep_http_request`, … |
 
-**Common `OmniRunnerHandler` entry points** (offline — grep `runtime.py` before adding):
+**Common `OfflineOmniClient` entry points** (offline — grep `runtime.py` before adding):
 
 | Area | High-level (`send_*_request` + assert) | Notes |
 |------|----------------------------------------|-------|
@@ -425,7 +426,7 @@ Changing `request_config` fields (`size`, `n`, `seed`, server flags) is **not**
 | Single-stage TTS (Coqui XTTS, MOSS-TTS-Nano, …) | `send_single_stage_tts_request`, `send_single_stage_tts_batch_request` | `prompt` + `additional_information` + `omni.generate` — **do not** duplicate in tests |
 | New model family | **Add** `send_<family>_request` here first | Then call from `tests/e2e/offline_inference/test_*.py` |
 
-L1 tests under `tests/entrypoints/` may use **FastAPI `TestClient`** or direct handler calls with mocks; they do **not** need `OpenAIClientHandler` / `OmniRunnerHandler`. **L2+ online and offline e2e** must use `runtime.py` `send_*` helpers.
+L1 tests under `tests/entrypoints/` may use **FastAPI `TestClient`** or direct handler calls with mocks; they do **not** need `OnlineOmniClient` / `OfflineOmniClient`. **L2+ online and offline e2e** must use `runtime.py` `send_*` helpers.
 
 #### Invalid parameter validation (`tests/dfx/reliability/invalid_param_test/`)
 
@@ -451,7 +452,7 @@ When the user asks for **invalid parameter validation**, **invalid request bodie
 | **Module markers** | `pytestmark = [pytest.mark.slow, pytest.mark.<omni or tts or diffusion>]` |
 | **Server fixture** | `_PARAMS` / `_QWEN3_TTS_SPEECH`-style list of `pytest.param(OmniServerParams(...), id="...", marks=hardware_marks(...))`; `@pytest.mark.parametrize("omni_server", _PARAMS, indirect=True)` |
 | **Hardware** | `hardware_marks(res={"cuda": "H100"})` for heavy diffusion/omni/video; `hardware_marks(res={"cuda": "L4"})` for smaller TTS models (must match weekly `-m "slow and L4"` step) |
-| **HTTP client** | **Low-level** `openai_client.send_*_http_request({..., "err_code": 400, "err_message": (...)} )` — **not** `send_*_request` (success path) |
+| **HTTP client** | **Low-level** `online_client.send_*_http_request({..., "err_code": 400, "err_message": (...)} )` — **not** `send_*_request` (success path) |
 | **Case shape** | Prefer **one `test_*` per route family** + `@pytest.mark.parametrize("body_spec, err_message", [...])` with stable `id=` per case; or dedicated `test_<route>_malformed_json` when not parametrized |
 | **Body helpers** | `_minimal_<endpoint>_json(omni_server)` / `_minimal_*_form_data()`; merge overrides with `body.update(body_spec)` |
 | **Known gaps** | `pytest.mark.skip(reason="…#3649")` as `_SKIP_ISSUE_3649` when server validation is not yet strict (mirror neighboring cases) |
@@ -484,13 +485,13 @@ When the user asks for **invalid parameter validation**, **invalid request bodie
 @pytest.mark.parametrize("omni_server", _PARAMS, indirect=True)
 def test_images_generations_invalid_requests(
     omni_server: OmniServer,
-    openai_client: OpenAIClientHandler,
+    online_client: OnlineOmniClient,
     body_spec: dict[str, object],
     err_message: str | tuple[str, ...],
 ) -> None:
     body = _minimal_images_gen_json(omni_server)
     body.update(body_spec)
-    openai_client.send_images_generations_http_request(
+    online_client.send_images_generations_http_request(
         {"json": body, "timeout": 300, "err_code": 400, "err_message": err_message}
     )
 ```
@@ -541,18 +542,18 @@ def send_images_generations_request(self, request_config: dict[str, Any], ...) -
 
 # tests/e2e/.../test_foo_expansion.py — one case per test_* (no case_id branching)
 @pytest.mark.parametrize("omni_server", [pytest.param(OmniServerParams(model=MODEL), marks=SINGLE_L4)], indirect=True)
-def test_foo_images_generations_default_1024(omni_server, openai_client) -> None:
-    openai_client.send_images_generations_request({"json": body, "timeout": 300})
+def test_foo_images_generations_default_1024(omni_server, online_client) -> None:
+    online_client.send_images_generations_request({"json": body, "timeout": 300})
 
 @pytest.mark.parametrize("omni_server", [pytest.param(OmniServerParams(model=MODEL), marks=SINGLE_L4)], indirect=True)
-def test_foo_images_generations_sizes_256_512_1024(omni_server, openai_client) -> None:
+def test_foo_images_generations_sizes_256_512_1024(omni_server, online_client) -> None:
     for size in ("256x256", "512x512", "1024x1024"):
-        openai_client.send_images_generations_request({"json": {**body, "size": size}, "timeout": 300})
+        online_client.send_images_generations_request({"json": {**body, "size": size}, "timeout": 300})
 
 # chat diffusion — separate test; send_diffusion_request bundles assert_diffusion_response
 @pytest.mark.parametrize("omni_server", [pytest.param(OmniServerParams(model=MODEL), marks=SINGLE_L4)], indirect=True)
-def test_foo_chat_completions_t2i_fallback(omni_server, openai_client) -> None:
-    openai_client.send_diffusion_request({...})
+def test_foo_chat_completions_t2i_fallback(omni_server, online_client) -> None:
+    online_client.send_diffusion_request({...})
 ```
 
 L1 tests under `tests/entrypoints/` may keep **minimal** asserts next to the handler under test when validating a single branch; still prefer `assertions.py` when the same JSON/media check appears in more than one file or level (e.g. L1 protocol test + L4 e2e).

.claude/skills/vllm-omni-test/references/test-routing.md

@@ -279,7 +279,7 @@ When adding or modifying tests, do not stop at “where the file lives” — al
 4. **API client + assert placement**: **General** validation → implement in `assertions.py`, call **inside** `send_*_request` in `runtime.py`; tests call **`send_*_request` only**. **Special** case validation → `assert_*` in `assertions.py`, called **in the test** after `send_*_request`. Low-level `send_*_http_request` is for negative/dfx tests (`err_code`), not ordinary L2+ success e2e.
 5. **Shared assertions**: logic in **`tests/helpers/assertions.py`**; general checks inside `send_*_request`; special checks only in the test. No `_assert_*` in `test_*.py`.
 6. **One case → one `test_*`**: function name reflects what is validated; no `if case_id == ...` mega-test merging multiple scenarios.
-7. **Fixture scope**: default **`omni_server` + `openai_client`** / **`omni_runner` + `omni_runner_handler`** (module). Use **`omni_server_function` + `openai_client_function`** / **`omni_runner_function` + `omni_runner_handler_function`** only when each `test_*` must spawn a fresh instance. Parametrize name must match fixture (`omni_server` vs `omni_server_function`).
+7. **Fixture scope**: default **`omni_server` + `online_client`** / **`omni_runner` + `offline_client`** (module). Use **`omni_server_function` + `online_client_function`** / **`omni_runner_function` + `offline_client_function`** only when each `test_*` must spawn a fresh instance. Parametrize name must match fixture (`omni_server` vs `omni_server_function`).
 8. **Type marker**: `omni`, `tts`, or `diffusion` on every model e2e module.
 9. **Diffusion L4 Function**: wire `*_expansion.py` into **X2I(&A&T)** or **X2V** **Function Test** in **`test-nightly.yml` only** — do not add `test-merge.yml` unless the user also requested L3.
 10. **Diffusion L4 Perf** (only when requested): add `tests/dfx/perf/tests/test_<slug>_vllm_omni.json` + **Perf Test · &lt;Model&gt;** step (artifact upload); not part of “L4 functional cases” by default.

.github/CODEOWNERS

@@ -79,8 +79,6 @@
 /vllm_omni/model_executor/models/ @tzhouam @gcanlin @Sy0307 @amy-why-3459 @linyueqian
 /vllm_omni/model_extras/ @tzhouam @gcanlin @princepride
 /vllm_omni/plugins/ @gcanlin @tjtanaa @xuechendi
-# Config override within model_executor — docs/design/module/vllm_omni_config.md
-/vllm_omni/model_executor/stage_configs/ @lishunyang12 @alex-jw-brooks
 
 # Execution platforms (hardware backends) — docs/design/module/execution_platforms.md
 /vllm_omni/platforms/ @gcanlin @tjtanaa @xuechendi

.github/ISSUE_TEMPLATE/400-bug-report.yml

@@ -83,7 +83,7 @@ body:
 
       omni = Omni(
           model="Qwen/Qwen-Image",
-          stage_configs_path="/path/to/stage_configs.yaml",
+          deploy_config="/path/to/deploy_config.yaml",
       )
 
       prompts = [{"prompt": "A scenic watercolor painting of a lighthouse at sunset"}]

.github/mergify.yml

@@ -45,7 +45,7 @@ pull_request_rules:
         > # For mypy (substitute "3.10" with the failing version if needed)
         > pre-commit run --hook-stage manual mypy-3.10
         > # For markdownlint
-        > pre-commit run --hook-stage manual markdownlint
+        > pre-commit run markdownlint-cli2
         > ```
         > </details>
 

.github/workflows/pre-commit.yml

@@ -24,9 +24,7 @@ jobs:
           python-version: "3.12"
           cache: 'pip'
       - uses: pre-commit/action@v3.0.1
-        # Don't run the pytest mark check against everything in the CI
-        # for now since a lot of files are still missing them. This check
-        # is used to help contributors remember to add the marks when they
-        # commit.
+        # Don't run historical-debt hooks against the whole tree in CI yet.
+        # They still run locally on changed files.
         env:
-          SKIP: check-test-ci-coverage
+          SKIP: check-test-ci-coverage,markdownlint-cli2,shellcheck,check-spdx-header,mypy-3.10

.gitignore

@@ -242,9 +242,9 @@ data/
 datasets/
 *.csv
 *.json
+!recipes/cosmos3/negative_prompt.json
 !tests/dfx/perf/tests/*.json
 !apps/ComfyUI-vLLM-Omni/example_workflows/*.json
-!tests/assets/soulxsinger/*.json
 *.jsonl
 *.parquet
 
@@ -254,6 +254,7 @@ outputs/
 results/
 generated/
 output_*/
+/tests/e2e/accuracy/artifacts/
 
 # Configuration overrides
 configs/local.yaml
@@ -277,3 +278,6 @@ tests/.ci_generated/
 
 # Local Lance precision-matching dump scripts (not part of the repo)
 lance_compare/
+
+# Linting binaries downloaded by local pre-commit hooks
+shellcheck*/

.markdownlint.yaml

@@ -0,0 +1,17 @@
+# Keep these rules aligned with vLLM so MkDocs Material syntax is tolerated.
+MD007:
+  indent: 4
+MD013: false
+# GitHub READMEs start with a centered logo, not an H1.
+MD041: false
+MD024:
+  siblings_only: true
+MD025:
+  # Allow front matter title to be different from the first heading in the document.
+  front_matter_title: ""
+MD031:
+  list_items: false
+MD033: false
+MD046: false
+MD052: false
+MD059: false

.pre-commit-config.yaml

@@ -24,7 +24,7 @@ repos:
     - id: ruff-format
 
   - repo: https://github.com/crate-ci/typos
-    rev: typos-dict-v0.13.13
+    rev: v1.49.0
     hooks:
       - id: typos
         # only for staged files
@@ -38,7 +38,19 @@ repos:
       - id: actionlint
         files: ^\.github/workflows/.*\.ya?ml$
 
+  - repo: https://github.com/DavidAnson/markdownlint-cli2
+    rev: v0.23.2
+    hooks:
+      - id: markdownlint-cli2
+        language_version: lts
+        args: [--fix]
+        files: (^docs/|^recipes/|(^|/)README\.md$|^CONTRIBUTING\.md$)
+        # Agent skill docs and CLAUDE.md are not user-facing docs/READMEs.
+        exclude: (^|/)CLAUDE\.md$|^\.claude/|^\.cursor/
 
+  # Keep `local` last so `suggestion` stays the final hook. Autofix repos
+  # (ruff, typos, markdownlint, trailing-whitespace, etc.) must run first;
+  # otherwise their "files were modified" output displaces the SKIP tip.
   - repo: local
     hooks:
     - id: signoff-commit
@@ -54,6 +66,45 @@ repos:
       verbose: true
       stages: [commit-msg]
 
+    # TODO: Use https://github.com/pre-commit/mirrors-mypy when mypy setup
+    # is less awkward. Local wrapper groups files so follow_imports=skip on
+    # tests does not leak into library code.
+    # mypy-3.10 runs on git commit (changed files). CI SKIPs it because
+    # --all-files is not clean yet. Extra versions are manual-stage only
+    # (not in default `pre-commit run --all-files`):
+    #   pre-commit run --hook-stage manual mypy-3.12
+    - id: mypy-3.10
+      name: Run mypy for Python 3.10
+      entry: python tools/pre_commit/mypy.py "3.10"
+      <<: &mypy_common
+        language: python
+        types_or: [python, pyi]
+        require_serial: true
+        additional_dependencies:
+          - mypy==1.11.1
+          - regex
+          - types-cachetools
+          - types-setuptools
+          - types-PyYAML
+          - types-requests
+          - types-torch
+          - pydantic
+    - id: mypy-3.11
+      name: Run mypy for Python 3.11
+      entry: python tools/pre_commit/mypy.py "3.11"
+      <<: *mypy_common
+      stages: [manual]
+    - id: mypy-3.12
+      name: Run mypy for Python 3.12
+      entry: python tools/pre_commit/mypy.py "3.12"
+      <<: *mypy_common
+      stages: [manual]
+    - id: mypy-3.13
+      name: Run mypy for Python 3.13
+      entry: python tools/pre_commit/mypy.py "3.13"
+      <<: *mypy_common
+      stages: [manual]
+
     - id: check-test-ci-coverage
       name: Ensure test files have pytestmarks
       entry: python tools/pre_commit/check_test_marks.py
@@ -66,12 +117,50 @@ repos:
       entry: python tools/pre_commit/check_tts_adapter.py
       language: python
       pass_filenames: false
-      files: ^(vllm_omni/entrypoints/openai/(serving_speech\.py|tts_adapters/__init__\.py)|tools/pre_commit/check_tts_adapter\.py)$
+      files: ^(vllm_omni/entrypoints/openai/serving_speech\.py|tools/pre_commit/check_tts_adapter\.py)$
       # The budget-is-now-slack reminder is printed on a PASSING run, and
       # pre-commit hides output from passing hooks unless the hook is verbose.
       verbose: true
 
-    # Keep `suggestion` last
+    - id: shellcheck
+      name: Lint shell scripts
+      # language: script breaks on Windows (shebang path loses slashes).
+      # Invoke the vLLM-style wrapper with bash so Linux CI and Win/Git Bash work.
+      entry: bash
+      args: [tools/pre_commit/shellcheck.sh]
+      language: system
+      types: [shell]
+
+    - id: check-spdx-header
+      name: Check SPDX headers
+      entry: python tools/pre_commit/check_spdx_header.py
+      language: python
+      types_or: [python, pyi, rust, proto, shell]
+      files: \.(py|pyi|rs|proto|sh)$
+
+    - id: check-forbidden-imports
+      name: Check forbidden imports
+      entry: python tools/pre_commit/check_forbidden_imports.py
+      language: python
+      types: [python]
+      additional_dependencies: [regex]
+
+    - id: check-torch-cuda-call
+      name: Prevent new torch.cuda API calls
+      entry: python tools/pre_commit/check_torch_cuda.py
+      language: python
+      types: [python]
+      additional_dependencies: [regex]
+
+    - id: check-buildkite
+      name: Validate Buildkite Pipelines
+      entry: python tools/pre_commit/check_buildkite.py
+      language: python
+      files: ^\.buildkite/.*\.ya?ml$
+      additional_dependencies: [pyyaml, check-jsonschema==0.38.0]
+
+    # Keep `suggestion` as the final hook. Run after autofix hooks so the
+    # bypass/SKIP tip is not displaced.
     - id: suggestion
       name: Suggestion
       entry: bash -c 'echo "To bypass all the pre-commit hooks, add --no-verify to git commit. To skip a specific hook, prefix the commit command with SKIP=<hook-id>."'
@@ -79,10 +168,3 @@ repos:
       verbose: true
       pass_filenames: false
     # Insert new entries above the `suggestion` entry
-
-    - id: check-pickle-imports
-      name: Prevent new pickle/cloudpickle imports
-      entry: python tools/pre_commit/check_pickle_imports.py
-      language: python
-      types: [python]
-      additional_dependencies: [regex]

CONTRIBUTING.md

@@ -1,5 +1,11 @@
 # Contributing to vLLM-Omni
 
-You may find information about contributing to vLLM-Omni on [Contributing](https://vllm-omni.readthedocs.io/en/latest/contributing/)
+You may find information about contributing to vLLM-Omni on [Contributing](https://vllm-omni.readthedocs.io/en/latest/contributing/).
+
+Local `pre-commit` gates (SPDX, forbidden imports, `torch.cuda`, shellcheck,
+mypy, test marks, TTS adapter ratchet, markdownlint, Buildkite schema, and
+how to extend allowlists) are documented in the
+[Linting](https://vllm-omni.readthedocs.io/en/latest/contributing/#linting)
+section of that guide.
 
 Before submitting a PR, run the [precheck-pr skill](https://github.com/vllm-project/vllm-omni/blob/main/.claude/skills/precheck-pr/SKILL.md) with the code agent for a self-check against project conventions.

README.md

@@ -12,10 +12,10 @@ Easy, fast, and cheap omni-modality model serving for everyone
 | <a href="https://vllm-omni.readthedocs.io/en/latest/"><b>Documentation</b></a> | <a href="https://deepwiki.com/vllm-project/vllm-omni"><b>DeepWiki</b></a> | <a href="https://discuss.vllm.ai"><b>User Forum</b></a> | <a href="https://slack.vllm.ai"><b>Developer Slack</b></a> | <a href="docs/assets/WeChat.jpg"><b>WeChat</b></a> | <a href="https://arxiv.org/abs/2602.02204"><b>Paper</b></a> | <a href="https://docs.google.com/presentation/d/1aPj0OGl_-ZVoib-Qne5dGDAlrRFB-PdHl6E-EE99g8E/edit?usp=sharing"><b>Slides</b></a> |
 </p>
 
-
 ---
 
 *Latest News* 🔥
+
 - [2026/08] [VeRL-Omni](https://github.com/verl-project/verl-omni) `v0.2.0` is released: faster diffusion RL powered by vLLM-Omni (request-level/step-wise batching with FA3), rebuilt Qwen3-Omni multimodal training (DPO & GSPO), plus LTX-2.3, Qwen-Image-Edit support and more. See the [release notes](https://github.com/verl-project/verl-omni/releases/tag/v0.2.0).
 - [2026/08] We released [0.26.0](https://github.com/vllm-project/vllm-omni/releases/tag/v0.26.0) - aligned with the vLLM 0.26 release line, featuring [MiniMax H3](recipes/MiniMaxAI/MiniMax-H3.md) joint video/audio generation, an experimental full-duplex realtime runtime for [MiniCPM-o 4.5](recipes/OpenBMB/MiniCPM-o-4_5.md), distributed layerwise diffusion offload, and broader model, hardware, streaming, TTS, and quantization support.
 - [2026/07] We released [0.24.0](https://github.com/vllm-project/vllm-omni/releases/tag/v0.24.0) - aligned with the vLLM 0.24 release line, expanding production-ready coverage across TTS, speech, diffusion, image/video generation, and robot-policy serving, with major Omni stage runtime refactoring, diffusion request-level batching, async output materialization, quantization/cache/memory improvements, and broad CUDA/ROCm/XPU/NPU support.
@@ -89,11 +89,18 @@ If you use vLLM-Omni for your research, please cite our [paper](https://arxiv.or
 ```
 
 ## Join the Community
+
 Feel free to ask questions, provide feedbacks and discuss with fellow users of vLLM-Omni in `#sig-omni` slack channel at [slack.vllm.ai](https://slack.vllm.ai) or vLLM user forum at [discuss.vllm.ai](https://discuss.vllm.ai).
 
 ## Star History
 
-[![Star History Chart](https://api.star-history.com/svg?repos=vllm-project/vllm-omni&type=date&legend=top-left)](https://www.star-history.com/#vllm-project/vllm-omni&type=date&legend=top-left)
+<a href="https://www.star-history.com/?repos=vllm-project%2Fvllm-omni&type=date&legend=top-left">
+  <picture>
+    <source media="(prefers-color-scheme: dark)" srcset="https://api.star-history.com/chart?repos=vllm-project/vllm-omni&type=date&theme=dark&legend=top-left&sealed_token=ExgLDZJoQEg27Zfhhut2LqN0GYO6Fw2PWLwPE6JYBUp2BgM3hmsYlwaIVopnUEfbRXidQ4nisumrTdKYydiKhy1SZXipw47qY2_tiUDhCpsPXeXtPuEVKVzBwKs3pw0tiHsJgtSfwXx5yjHXck0Y2SblzFWeJYCkTe1WLGTbUAOIETjXXQJjyCGZvKz5" />
+    <source media="(prefers-color-scheme: light)" srcset="https://api.star-history.com/chart?repos=vllm-project/vllm-omni&type=date&legend=top-left&sealed_token=ExgLDZJoQEg27Zfhhut2LqN0GYO6Fw2PWLwPE6JYBUp2BgM3hmsYlwaIVopnUEfbRXidQ4nisumrTdKYydiKhy1SZXipw47qY2_tiUDhCpsPXeXtPuEVKVzBwKs3pw0tiHsJgtSfwXx5yjHXck0Y2SblzFWeJYCkTe1WLGTbUAOIETjXXQJjyCGZvKz5" />
+    <img alt="Star History Chart" src="https://api.star-history.com/chart?repos=vllm-project/vllm-omni&type=date&legend=top-left&sealed_token=ExgLDZJoQEg27Zfhhut2LqN0GYO6Fw2PWLwPE6JYBUp2BgM3hmsYlwaIVopnUEfbRXidQ4nisumrTdKYydiKhy1SZXipw47qY2_tiUDhCpsPXeXtPuEVKVzBwKs3pw0tiHsJgtSfwXx5yjHXck0Y2SblzFWeJYCkTe1WLGTbUAOIETjXXQJjyCGZvKz5" />
+  </picture>
+</a>
 
 ## License
 

benchmarks/tts/README.md

@@ -18,8 +18,8 @@ vllm serve Qwen/Qwen3-TTS-12Hz-1.7B-Base --omni --port 8000
 ```
 
 The server auto-loads its Deploy YAML from `vllm_omni/deploy/qwen3_tts.yaml`
-(Pipeline + Deploy schema introduced in #2383). No `--stage-configs-path` or
-`--deploy-config` flag is needed for any registered model.
+(Pipeline + Deploy schema introduced in #2383). No explicit `--deploy-config`
+flag is needed for a registered model.
 
 ### 2. Run the benchmark (`vllm bench serve --omni`)
 

docs/.nav.yml

@@ -43,6 +43,7 @@ nav:
     - Overview: features/README.md
     - Runtime and Stage Execution:
       - Execution Modes and Streaming: user_guide/diffusion/execution_modes.md
+      - Prefill-Decode Disaggregation (experimental): features/pd_disaggregation.md
       - Sleep Mode: features/sleep_mode.md
     - Quantization:
       - Overview: user_guide/quantization/overview.md
@@ -60,7 +61,11 @@ nav:
       - Overview and Compatibility:
         - Overview: user_guide/diffusion_features.md
         - Feature Compatibility: user_guide/feature_compatibility.md
-      - CPU Offloading: user_guide/diffusion/cpu_offload.md
+      - CPU Offloading:
+        - Overview: user_guide/diffusion/cpu_offload.md
+        - Model-Level: user_guide/diffusion/offloader/cpu_offload.md
+        - Layerwise: user_guide/diffusion/offloader/layerwise_offload.md
+        - Distributed Layerwise: user_guide/diffusion/offloader/distributed_layerwise_offload.md
       - Cache Acceleration:
         - TeaCache: user_guide/diffusion/cache_acceleration/teacache.md
         - Cache-DiT: user_guide/diffusion/cache_acceleration/cache_dit.md
@@ -72,7 +77,13 @@ nav:
         - Sequence Parallel: user_guide/diffusion/parallelism/sequence_parallel.md
         - Tensor Parallel: user_guide/diffusion/parallelism/tensor_parallel.md
         - VAE Parallelism: user_guide/diffusion/parallelism/vae_parallelism.md
-      - Attention Backends: user_guide/diffusion/attention_backends.md
+      - Attention Backends:
+        - Overview: user_guide/diffusion/attention_backends.md
+        - Dense Backends: user_guide/diffusion/attention_backends/dense_backends.md
+        - TRTLLM: user_guide/diffusion/attention_backends/trtllm.md
+        - SageAttention: user_guide/diffusion/attention_backends/sage.md
+        - Hugging Face Hub: user_guide/diffusion/attention_backends/huggingface_hub.md
+        - RainFusion: user_guide/diffusion/attention_backends/rainfusion.md
       - Regional Compilation: user_guide/diffusion/regional_compilation.md
       - Frame Interpolation: user_guide/diffusion/frame_interpolation.md
       - Startup and Loading: user_guide/diffusion/startup_and_loading.md
@@ -102,10 +113,12 @@ nav:
     - Feature Design:
       - Runtime and Stage Execution:
         - Disaggregated Inference: design/feature/disaggregated_inference.md
+        - Host Weight Runtime: design/feature/host_weight_runtime.md
         - Async Chunk: design/feature/async_chunk.md
         - Async Diffusion Output: design/feature/async_diffusion_output.md
         - Async Omni Output Materialization: design/feature/omni_async_output_materialization.md
         - Automatic Prefix Caching: design/feature/prefix_caching.md
+        - Realtime AR-Diffusion: design/feature/realtime_ar_diffusion.md
       - Communication:
         - OmniConnector Implementations:
           - Mooncake Store Connector: design/feature/omni_connectors/mooncake_store_connector.md
@@ -125,11 +138,16 @@ nav:
           - Tensor Parallel: design/feature/tensor_parallel.md
           - VAE Patch Parallelism: design/feature/vae_parallel.md
         - Attention Optimization:
+          - Backend Selection: design/feature/attention_backend_selection.md
           - Skip-Softmax: design/feature/skip_softmax.md
+        - CPU Offloading:
+          - Overview: design/feature/offloader/README.md
+          - Model-Level: design/feature/offloader/cpu_offload.md
+          - Layerwise: design/feature/offloader/layerwise_offload.md
+          - Distributed Layerwise: design/feature/offloader/distributed_layerwise_offload.md
         - Cache-DiT: design/feature/cache_dit.md
         - TeaCache: design/feature/teacache.md
         - Diffusion Continuous Batching: design/feature/diffusion_continuous_batching.md
-        - Distributed Layerwise Offload: design/feature/distributed_layerwise_offload.md
     - Infrastructure and Performance:
       - Prometheus Metrics: design/metrics.md
       - Speech Generation Performance Optimizations: design/qwen3_omni_tts_performance_optimization.md
@@ -152,6 +170,7 @@ nav:
         - Offloader: design/module/diffusion/offloader.md
       - Execution Platforms: design/module/execution_platforms.md
       - Cache Management: design/module/cache_management.md
+      - Host Weight Runtime: design/module/host_weight_runtime.md
       - Quantization: design/module/quantization.md
       - Observability: design/module/observability.md
       - Profiling: design/module/profiling.md

docs/api/README.md

@@ -114,10 +114,6 @@ Configuration classes.
 - [vllm_omni.model_executor.models.qwen3_tts.tokenizer_25hz.configuration_qwen3_tts_tokenizer_v1.Qwen3TTSTokenizerV1DecoderDiTConfig][]
 - [vllm_omni.model_executor.models.qwen3_tts.tokenizer_25hz.configuration_qwen3_tts_tokenizer_v1.Qwen3TTSTokenizerV1EncoderConfig][]
 - [vllm_omni.transformers_utils.configs.cosyvoice3.CosyVoice3Config][]
-- [vllm_omni.transformers_utils.configs.mammoth_moda2.Mammothmoda2Config][]
-- [vllm_omni.transformers_utils.configs.mammoth_moda2.Mammothmoda2Qwen2_5_VLConfig][]
-- [vllm_omni.transformers_utils.configs.mammoth_moda2.Mammothmoda2Qwen2_5_VLTextConfig][]
-- [vllm_omni.transformers_utils.configs.mammoth_moda2.Mammothmoda2Qwen2_5_VLVisionConfig][]
 - [vllm_omni.transformers_utils.configs.omnivoice.OmniVoiceConfig][]
 - [vllm_omni.transformers_utils.configs.voxcpm2.VoxCPM2Config][]
 

docs/configuration/README.md

@@ -4,20 +4,17 @@ This section lists the most common options for running vLLM-Omni.
 
 For options within a vLLM Engine, please refer to the [vLLM 0.26 configuration guide](https://docs.vllm.ai/en/v0.26.0/configuration/index.html).
 
-Currently, the main options are maintained by stage configs for each model.
+Each model defines fixed topology in a registered `PipelineConfig` and runtime
+overrides in a deploy YAML.
 
 For a specific example, see the [Qwen2.5-Omni deploy config](gh-file:vllm_omni/deploy/qwen2_5_omni.yaml). The matching frozen pipeline topology lives at [vllm_omni/model_executor/models/qwen2_5_omni/pipeline.py](gh-file:vllm_omni/model_executor/models/qwen2_5_omni/pipeline.py).
 
-For introduction, please check [Introduction for stage config](./stage_configs.md)
+For an introduction, see [Pipeline and deploy configurations](./stage_configs.md).
 
 ## Memory Configuration
 
 - **[GPU Memory Calculation and Configuration](./gpu_memory_utilization.md)** - Guide on how to calculate memory requirements and set up `gpu_memory_utilization` for optimal performance
 
-## Multi-Stage Recipes
-
-- **[Prefill-Decode Disaggregation](./pd_disaggregation.md)** - How to derive a PD-aware Qwen3-Omni stage config from the default config without introducing another bundled YAML
-
 ## Optimization Features
 
 - **[Diffusion Features Overview](../user_guide/diffusion_features.md)** - Complete overview of all diffusion model features and supported models

docs/configuration/composable_parallel.md

@@ -1,6 +1,6 @@
 # Composable parallel strategies
 
-In vLLM-Omni, the `composable_parallel` system layers a declarative *parallel strategy* on top of the existing stage configs. A strategy YAML maps each pipeline stage (by its `model_stage` name) to a stack of axis declarations — one per parallelism dimension — that the engine translates into per-stage parallel sizing (`tensor_parallel_size`, `data_parallel_size`, `pipeline_parallel_size`, expert-parallel toggle, replica count) and, when relevant, an orchestrator-wide load-balancer policy.
+In vLLM-Omni, the `composable_parallel` system layers a declarative *parallel strategy* on top of the registered pipeline and deploy config. A strategy YAML maps each pipeline stage (by its `model_stage` name) to a stack of axis declarations — one per parallelism dimension — that the engine translates into per-stage parallel sizing (`tensor_parallel_size`, `data_parallel_size`, `pipeline_parallel_size`, expert-parallel toggle, replica count) and, when relevant, an orchestrator-wide load-balancer policy.
 
 Strategies are intentionally narrow: each one only writes the axes it declares, leaving every other engine arg untouched. This makes a strategy file a small, reusable overlay you can swap independently of the deploy YAML (see `configuration/stage_configs.md`).
 
@@ -195,7 +195,7 @@ Because `omni_lb_policy` is read once at `AsyncOmniEngine` construction — not
 
 Default: no default — required for every entry under `strategies:`.
 
-Strategy keys MUST be `model_stage` *names* (strings), the same labels stage configs use for `engine_args.model_stage` (see [`model_stage` in `stage_configs.md`](stage_configs.md#engine_argsmodel_stage)). Integer `stage_id` values are not accepted; the resolver matches on `model_stage` exactly (`_resolve_stage` in `vllm_omni/config/composable_parallel/apply.py`).
+Strategy keys MUST be `model_stage` *names* (strings), the same labels defined by [`StagePipelineConfig.model_stage`](stage_configs.md#pipeline-configuration). Integer `stage_id` values are not accepted; the resolver matches on `model_stage` exactly (`_resolve_stage` in `vllm_omni/config/composable_parallel/apply.py`).
 
 ## Common pitfalls
 

docs/configuration/gpu_memory_utilization.md

@@ -80,17 +80,13 @@ When multiple stages share the same GPU, you must ensure the sum of their `gpu_m
 
 **Example: Two stages on GPU 0**
 ```yaml
-stage_args:
+stages:
   - stage_id: 0
-    runtime:
     devices: "0"
-    engine_args:
     gpu_memory_utilization: 0.6  # Uses 60% of GPU 0
 
   - stage_id: 1
-    runtime:
     devices: "0"
-    engine_args:
     gpu_memory_utilization: 0.3  # Uses 30% of GPU 0
     # Total: 90% of GPU 0 (safe, leaves 10% buffer)
 ```
@@ -103,11 +99,9 @@ When using `tensor_parallel_size > 1`, the model is split across multiple GPUs,
 
 **Example: 2-way tensor parallelism**
 ```yaml
-stage_args:
+stages:
   - stage_id: 0
-    runtime:
     devices: "0,1"  # Uses both GPUs
-    engine_args:
     tensor_parallel_size: 2
     gpu_memory_utilization: 0.6  # 60% per GPU
     # Model is split, so each GPU uses ~30% of model memory
@@ -118,27 +112,25 @@ stage_args:
 ### Qwen3-Omni-MoE on 2x H100-80GB
 
 ```yaml
-stage_args:
+base_config: /path/to/vllm_omni/deploy/qwen3_omni_moe.yaml
+
+stages:
   - stage_id: 0  # Thinker stage with TP=2
-    runtime:
     devices: "0,1"
-    engine_args:
     tensor_parallel_size: 2
     gpu_memory_utilization: 0.6  # 48GB per GPU
 
   - stage_id: 1  # Talker stage
-    runtime:
     devices: "1"
-    engine_args:
     gpu_memory_utilization: 0.3  # 24GB on GPU 1
 
   - stage_id: 2  # Code2Wav stage
-    runtime:
     devices: "0"
-    engine_args:
     gpu_memory_utilization: 0.1  # 8GB on GPU 0
 ```
-**Note:** In this configuration, stages 0 and 2 share GPU 0, but they run at different times in the pipeline, so their memory usage doesn't overlap.
+**Note:** Stage 0 uses GPUs 0 and 1, so the combined utilization is `0.7` on
+GPU 0 and `0.9` on GPU 1. Keep the sum of all resident stages below `1.0` on
+each device.
 
 ## Troubleshooting
 

docs/configuration/pd_disaggregation.md

@@ -1,171 +0,0 @@
-# Prefill-Decode (PD) Disaggregation
-
-PD disaggregation splits the Qwen3-Omni thinker into separate prefill and decode
-stages so prompt processing and token generation can run on different workers.
-
-This is documented as a stage-config recipe instead of a bundled YAML because the
-deployment-specific values usually change per environment:
-
-- GPU placement
-- `tensor_parallel_size`
-- connector backend and connector ports
-- connector IPs or bootstrap addresses
-
-Start from the [default Qwen3-Omni stage config](gh-file:vllm_omni/deploy/qwen3_omni_moe.yaml)
-and copy it to your own file, for example `qwen3_omni_pd.yaml`. Then apply the
-changes below.
-
-## Requirements
-
-- 3+ GPUs for a basic layout: prefill, decode, and talker+code2wav
-- A KV connector supported by vLLM, such as `MooncakeConnector`
-- Matching `tensor_parallel_size` on the prefill and decode thinker stages
-
-## 1. Split the thinker into prefill and decode stages
-
-Replace the original thinker stage with two stages:
-
-```yaml
-stage_args:
-  - stage_id: 0
-    stage_type: llm
-    is_prefill_only: true
-    runtime:
-      devices: "0"
-    engine_args:
-      max_num_seqs: 16
-      model_stage: thinker
-      model_arch: Qwen3OmniMoeForConditionalGeneration
-      worker_type: ar
-      scheduler_cls: vllm_omni.core.sched.omni_ar_scheduler.OmniARScheduler
-      gpu_memory_utilization: 0.9
-      enforce_eager: true
-      trust_remote_code: true
-      engine_output_type: latent
-      distributed_executor_backend: "mp"
-      enable_prefix_caching: false
-      max_num_batched_tokens: 32768
-      hf_config_name: thinker_config
-      tensor_parallel_size: 1
-      kv_transfer_config:
-        kv_connector: "MooncakeConnector"
-        kv_role: "kv_producer"
-        kv_rank: 0
-        kv_parallel_size: 2
-        kv_connector_extra_config:
-          mooncake_bootstrap_port: 25201
-    final_output: false
-    is_comprehension: true
-    default_sampling_params:
-      temperature: 0.4
-      top_p: 0.9
-      top_k: 1
-      max_tokens: 2048
-      seed: 42
-      detokenize: True
-      repetition_penalty: 1.05
-
-  - stage_id: 1
-    stage_type: llm
-    is_decode_only: true
-    runtime:
-      devices: "1"
-    engine_args:
-      max_num_seqs: 64
-      model_stage: thinker
-      model_arch: Qwen3OmniMoeForConditionalGeneration
-      worker_type: ar
-      scheduler_cls: vllm_omni.core.sched.omni_ar_scheduler.OmniARScheduler
-      gpu_memory_utilization: 0.9
-      enforce_eager: true
-      trust_remote_code: true
-      engine_output_type: latent
-      distributed_executor_backend: "mp"
-      enable_prefix_caching: false
-      max_num_batched_tokens: 32768
-      hf_config_name: thinker_config
-      tensor_parallel_size: 1
-      kv_transfer_config:
-        kv_connector: "MooncakeConnector"
-        kv_role: "kv_consumer"
-        kv_rank: 1
-        kv_parallel_size: 2
-        kv_connector_extra_config:
-          mooncake_bootstrap_port: 25202
-    engine_input_source: [0]
-    final_output: true
-    final_output_type: text
-    is_comprehension: true
-    default_sampling_params:
-      temperature: 0.4
-      top_p: 0.9
-      top_k: 1
-      max_tokens: 2048
-      seed: 42
-      detokenize: True
-      repetition_penalty: 1.05
-```
-
-Notes:
-
-- `is_prefill_only: true` marks the thinker stage that only saves KV.
-- `is_decode_only: true` marks the thinker stage that resumes from remote KV.
-- `kv_transfer_config` is required on both stages.
-- The orchestrator forces the prefill stage to run with `max_tokens=1`, so the
-  prefill side only processes the prompt and exports KV.
-
-## 2. Shift the downstream stages by one index
-
-After inserting the extra thinker stage, renumber the remaining stages:
-
-```yaml
-  - stage_id: 2
-    runtime:
-      devices: "2"
-    engine_input_source: [1]
-    sync_process_input_func: vllm_omni.model_executor.stage_input_processors.qwen3_omni.thinker2talker_token_only
-
-  - stage_id: 3
-    runtime:
-      devices: "2"
-    engine_args:
-      max_num_seqs: 1
-    engine_input_source: [2]
-    sync_process_input_func: vllm_omni.model_executor.stage_input_processors.qwen3_omni.talker2code2wav_token_only
-```
-
-Compared with the default Qwen3-Omni config:
-
-- the talker becomes stage `2` instead of stage `1`
-- the code2wav stage becomes stage `3` instead of stage `2`
-- the talker now reads from decode stage `1`
-
-## 3. Add runtime edges for the four-stage pipeline
-
-```yaml
-runtime:
-  enabled: true
-  edges:
-    - from: 0
-      to: 1
-    - from: 1
-      to: 2
-    - from: 2
-      to: 3
-```
-
-## 4. Launch with your custom config
-
-```bash
-vllm serve Qwen/Qwen3-Omni-30B-A3B-Instruct --omni --port 8091 \
-    --deploy-config /path/to/qwen3_omni_pd.yaml
-```
-
-## Operational Notes
-
-- `MooncakeConnector` does not support heterogeneous TP sizes across the PD
-  pair. Keep prefill and decode at the same `tensor_parallel_size`.
-- If the thinker requires TP=2, both thinker stages must use TP=2 and be given
-  separate GPU sets, for example `"0,1"` for prefill and `"2,3"` for decode.
-- Choose connector ports and addresses that match your deployment. The values
-  shown above are examples only.

docs/configuration/stage_configs.md

@@ -3,7 +3,46 @@
 In vLLM-Omni, a model's `PipelineConfig` defines its fixed stage topology, while a deploy configuration controls how those stages run.
 
 !!! note
-    Default deploy config YAMLs (for example, `vllm_omni/deploy/qwen2_5_omni.yaml`, `vllm_omni/deploy/qwen3_omni_moe.yaml`, and `vllm_omni/deploy/qwen3_tts.yaml`) are bundled and loaded automatically when `--deploy-config` is omitted. The model registry resolves the right pipeline and deploy YAML by `model_type`.
+    Default deploy config YAMLs (for example, `vllm_omni/deploy/qwen2_5_omni.yaml`, `vllm_omni/deploy/qwen3_omni_moe.yaml`, and `vllm_omni/deploy/qwen3_tts.yaml`) are bundled and loaded automatically when `--deploy-config` is omitted. The resolved pipeline selects its default through `default_deploy_config_name`.
+
+## Pipeline configuration
+
+`PipelineConfig` and its `StagePipelineConfig` entries are Python definitions
+owned by the model implementation and registered in `OMNI_PIPELINES`. They
+describe fixed model topology and are not accepted in deploy YAMLs.
+
+Common `PipelineConfig` fields include:
+
+| Field | Description |
+|-------|-------------|
+| `model_type` | Pipeline identifier used during model and config resolution. |
+| `default_deploy_config_name` | Bundled deploy YAML loaded when the user does not pass `deploy_config`. |
+| `model_arch` | Default Hugging Face architecture for the pipeline. |
+| `hf_architectures` | Architecture names used to identify checkpoints whose `model_type` is shared. |
+| `hf_config_predicate` | Optional predicate used to select between pipelines with otherwise identical HF metadata. |
+| `diffusers_class_name` | Diffusers `_class_name` used to identify Diffusers-style repositories. |
+| `stages` | Ordered tuple of fixed `StagePipelineConfig` definitions. |
+
+Common `StagePipelineConfig` fields include:
+
+| Field | Description |
+|-------|-------------|
+| `stage_id` | Stable stage identifier. |
+| `model_stage` | Logical stage name used by runtime and strategy resolution. |
+| `execution_type` | `LLM_AR`, `LLM_GENERATION`, or `DIFFUSION`. |
+| `input_sources` | Upstream stage IDs that provide this stage's inputs. |
+| `final_output` / `final_output_type` | Whether the stage produces a user-visible output and its modality. |
+| `owns_tokenizer` | Whether this stage owns the pipeline tokenizer. |
+| `model_arch`, `hf_config_name` | Stage-specific model architecture and nested HF config selector. |
+| `engine_output_type` | Runtime output representation such as `text`, `latent`, or `audio`. |
+| `custom_process_input_func` | Processor applied to this stage's incoming payload. |
+| `custom_process_next_stage_input_func` | Processor used for full-payload handoff to the next stage. |
+| `async_chunk_process_next_stage_input_func` | Processor used for async chunk handoff. |
+| `sampling_constraints` | Model-owned sampling constraints that deploy defaults cannot override. |
+
+To add or change topology, define and register a new pipeline variant. Use
+deploy YAML only for runtime placement, resource sizing, connectors, and other
+deployment overrides.
 
 ## Deploy configuration schema
 
@@ -13,42 +52,53 @@ The new deploy schema lives under `vllm_omni/deploy/` and is paired with a froze
 |-------|------|----------|---------|-------------|
 | `base_config` | str (path) | optional | — | Overlay parent (relative or absolute). `stages:` / `platforms:` deep-merged by stage_id; other scalars overlay-wins. Intended for user-authored overlays; prod yamls stay flat. |
 | `async_chunk` | bool | optional | `true` | Enable chunked streaming between stages. Pin to `false` if the pipeline runs end-to-end. |
+| `session_mode` | str | optional | `"turn"` | Session behavior. Use `"duplex"` only with a pipeline that enables duplex control. |
+| `active_stream_window` | int | optional | `0` | Number of active downstream stream slots; `0` preserves all-stream cycling. |
+| `duplex_session` | dict | optional | runtime defaults | Full-duplex session lifecycle, buffering, replay, and capacity limits. |
 | `connectors` | dict | optional | `null` | Named connector specs (`{name, extra}`). Referenced by each stage's `input_connectors` / `output_connectors`. See [Connector schema](#connector-schema). |
 | `edges` | list | optional | `null` | Explicit edge list for the KV transfer graph. Auto-derived from stage inputs if omitted. |
-| `stages` | list | required | — | Per-stage engine args + wiring (see [Stage fields](#stage-fields)). |
+| `stages` | list | optional | `[]` | Per-stage runtime overrides matched by `stage_id`. Pipeline stages are still created from `PipelineConfig` when this list is empty. |
 | `platforms` | dict | optional | `null` | Keyed by `npu` / `rocm` / `xpu`, each contains a `stages:` list with per-platform overrides applied on top of the CUDA defaults. |
 | `pipeline` | str | optional | `null` | Override the auto-detected pipeline registry key (used for structural variants like `qwen2_5_omni_thinker_only`). |
-| `trust_remote_code` | bool | optional | `true` | **Pipeline-wide.** Trust HF remote code on model load; applies to every stage. |
+| `trust_remote_code` | bool \| null | optional | `null` | **Pipeline-wide.** Trust HF remote code on model load; applies to every stage when specified. |
 | `distributed_executor_backend` | str \| null | optional | `null` | **Pipeline-wide.** Distributed executor backend forwarded to vLLM (`"mp"`, `"ray"`, `"external_launcher"`). If omitted, vLLM auto-selects backend from runtime topology. |
 | `dtype` | str \| null | optional | `null` | **Pipeline-wide.** Model dtype for every stage. |
 | `quantization` | str \| null | optional | `null` | **Pipeline-wide.** Quantization method for every stage. |
-| `enable_prefix_caching` | bool | optional | `false` | **Pipeline-wide.** Prefix cache toggle applied to every stage. |
+| `enable_prefix_caching` | bool \| null | optional | `null` | **Pipeline-wide.** Prefix cache toggle applied to every stage when specified. |
 | `enable_chunked_prefill` | bool \| null | optional | `null` | **Pipeline-wide.** Chunked prefill toggle applied to every stage. |
-| `data_parallel_size` | int | optional | `1` | **Pipeline-wide.** DP degree for every stage. |
-| `pipeline_parallel_size` | int | optional | `1` | **Pipeline-wide.** PP degree for every stage. |
+| `data_parallel_size` | int \| null | optional | `null` | **Pipeline-wide.** DP degree for every stage. |
+| `pipeline_parallel_size` | int \| null | optional | `null` | **Pipeline-wide.** PP degree for every stage. |
+| `custom_voice_dir` | str \| null | optional | `null` | **Pipeline-wide.** Directory containing custom voice profiles for supported TTS models. |
 
-Note: for diffusion path, `distributed_executor_backend` currently defaults to
-`mp`, and `ray` / `external_launcher` are not fully supported yet.
+For fields whose deploy default is `null`, the deploy layer contributes no
+override. The effective value may still come from a platform section, an
+explicit CLI or stage override, or the downstream vLLM engine default.
+
+Note: for the diffusion path, an omitted `distributed_executor_backend` selects
+`uni` on a single GPU (in-process worker, no MessageQueue / `/dev/shm` output
+segments) and `mp` when `num_gpus > 1`. Set `mp` explicitly to keep a worker
+subprocess on one GPU. `ray` / `external_launcher` are not fully supported yet.
 
 ### Stage fields
 
-Each entry under `stages:` accepts any `StageDeployConfig` field directly (no nested `engine_args:`). Only fields whose value legitimately varies across stages live here; pipeline-wide settings (trust_remote_code, distributed_executor_backend, dtype, quantization, prefix/chunked prefill, DP/PP sizes) are declared at the top level and applied to every stage. Unknown keys fall through to `engine_extras:` and are forwarded to the engine.
+Each entry under `stages:` accepts any `StageDeployConfig` field directly (no nested `engine_args:`). Only fields whose value legitimately varies across stages live here; pipeline-wide settings (trust_remote_code, distributed_executor_backend, dtype, quantization, prefix/chunked prefill, DP/PP sizes) are declared at the top level and applied to every stage. Unknown keys fall through to `engine_extras:` and are forwarded to the engine. Frequently used fields are listed below; the source-of-truth schema is `StageDeployConfig` in `vllm_omni/config/stage_config.py`.
 
 | Field | Type | Required | Default | Description |
 |-------|------|----------|---------|-------------|
 | `stage_id` | int | required | — | Stage identity; matched against `PipelineConfig.stages[*].stage_id`. |
-| `max_num_seqs` | int | optional | `64` | Max concurrent sequences per stage. |
-| `gpu_memory_utilization` | float | optional | `0.9` | Per-stage memory budget. |
-| `tensor_parallel_size` | int | optional | `1` | TP degree for this stage. |
-| `enforce_eager` | bool | optional | `false` | Disable CUDA graphs. |
-| `max_num_batched_tokens` | int | optional | `32768` | Prefill budget. |
-| `max_model_len` | int \| null | optional | `null` | Per-stage context length (auto-sets `VLLM_ALLOW_LONG_MAX_MODEL_LEN=1` when larger than HF default). |
+| `max_num_seqs` | int \| null | optional | `null` | Max concurrent sequences per stage. |
+| `gpu_memory_utilization` | float \| null | optional | `null` | Per-stage total memory target; used for automatic KV-cache sizing. |
+| `kv_cache_memory_bytes` | int \| null | optional | `null` | Explicit per-rank KV-cache budget in bytes; overrides automatic sizing. |
+| `tensor_parallel_size` | int \| null | optional | `null` | TP degree for this stage. |
+| `enforce_eager` | bool \| null | optional | `null` | Disable CUDA graphs. |
+| `max_num_batched_tokens` | int \| null | optional | `null` | Per-stage prefill/token budget; also contributes to the native maximum in-flight token limit. |
+| `max_model_len` | int \| null | optional | `null` | Per-sequence context or KV length; `-1` enables native cache-capacity auto-fitting, while values above the HF default auto-set `VLLM_ALLOW_LONG_MAX_MODEL_LEN=1`. |
 | `async_scheduling` | bool \| null | optional | `null` | Per-stage async scheduling toggle. |
-| `devices` | str | optional | `"0"` | `CUDA_VISIBLE_DEVICES`-style device list. |
+| `devices` | str \| null | optional | `null` | Device list assigned to this stage. |
 | `output_connectors` | dict \| null | optional | `null` | Keyed by `to_stage_<n>`; values are names registered under top-level `connectors:`. |
 | `input_connectors` | dict \| null | optional | `null` | Keyed by `from_stage_<n>`; values are names registered under top-level `connectors:`. |
 | `default_sampling_params` | dict \| null | optional | `null` | Baseline sampling params. Deep-merged with pipeline `sampling_constraints` (pipeline wins). |
-| `engine_extras` | dict | optional | `{}` | Catch-all for keys not listed above; deep-merged across overlays. Also carries per-stage overrides of pipeline-wide settings (e.g. stage-specific `dtype`). |
+| `engine_extras` | dict | optional | `{}` | Catch-all for engine fields not listed above; deep-merged across overlays and forwarded to the stage engine. |
 
 ### Connector schema
 
@@ -60,7 +110,7 @@ connectors:
     name: <ConnectorClassName>     # required — class registered in vllm_omni.distributed
     extra:                         # optional — forwarded to the connector's __init__
       <key>: <value>
-      ...
+      # Additional connector-specific options
 ```
 
 | Connector class | Use case | `extra` keys |
@@ -235,288 +285,3 @@ vllm serve Qwen/Qwen2.5-Omni-7B --omni --port 8091 --deploy-config /path/to/depl
 
 !!! important
     We are actively iterating on the definition of deployment configurations, and we welcome feedback from users and developers.
-
-## Legacy stage configuration reference
-
-The remainder of this page documents the deprecated `stage_args` schema for models that have not migrated to deploy configurations. The following historical Qwen2.5-Omni configuration illustrates that schema; current Qwen2.5-Omni deployments should use `vllm_omni/deploy/qwen2_5_omni.yaml` instead.
-
-```python
-# Historical stage configuration for Qwen2.5-Omni.
-stage_args:
-  - stage_id: 0 # mark the unique id for each stage
-    runtime: # The disaggregated configuration
-      process: true  # Run this stage in a separate process
-      devices: "0" # Logical device index for this stage (mapped through CUDA_VISIBLE_DEVICES / ASCEND_RT_VISIBLE_DEVICES if set)
-    engine_args: # Engine arguments for a certain engine
-      model_stage: thinker
-      max_num_seqs: 1
-      model_arch: Qwen2_5OmniForConditionalGeneration # The model implementation registered in model_executor/models/registry.py
-      worker_type: ar # The specific worker used
-      scheduler_cls: vllm_omni.core.sched.omni_ar_scheduler.OmniARScheduler # The specific scehduler used
-      gpu_memory_utilization: 0.8 # The gpu memory allocation for the stage within a single chip
-      enforce_eager: true  # Now we only support eager mode
-      trust_remote_code: true # Needed by huggingface config parsing
-      engine_output_type: latent  # It claims that the stage will input latent hiddenstates besides token ids
-      enable_prefix_caching: false # For request with hiddenstates output, the prefix caching is not supported now
-    is_comprehension: true # If the stage is a text or multimodal comprehension module. If it is, the AsyncOmni will use its tokenizer as default
-    final_output: true # If the stage has output as part of final outputs. If it is false, which means that the stage only works as a intermediate role.
-    final_output_type: text # What is the final output type. It can be text and audio now.
-    default_sampling_params: # sampling parameters for the stage. Their meaning aligns with vLLM.
-      temperature: 0.0
-      top_p: 1.0
-      top_k: -1
-      max_tokens: 2048
-      seed: 42
-      detokenize: True
-      repetition_penalty: 1.1
-  - stage_id: 1
-    runtime:
-      process: true
-      devices: "1"
-    engine_args:
-      model_stage: talker
-      max_num_seqs: 3
-      model_arch: Qwen2_5OmniForConditionalGeneration
-      worker_type: ar
-      scheduler_cls: vllm_omni.core.sched.omni_ar_scheduler.OmniARScheduler
-      gpu_memory_utilization: 0.8
-      enforce_eager: true
-      trust_remote_code: true
-      enable_prefix_caching: false
-      engine_output_type: latent
-    engine_input_source: [0]
-    default_sampling_params:
-      temperature: 0.9
-      top_p: 0.8
-      top_k: 40
-      max_tokens: 2048
-      seed: 42
-      detokenize: True
-      repetition_penalty: 1.05
-      stop_token_ids: [8294]
-  - stage_id: 2
-    runtime:
-      process: true
-      devices: "0"            # Example: use a different GPU than the previous stage; use "0" if single GPU
-    engine_args:
-      model_stage: code2wav
-      max_num_seqs: 1
-      model_arch: Qwen2_5OmniForConditionalGeneration
-      worker_type: generation
-      scheduler_cls: vllm_omni.core.sched.omni_generation_scheduler.OmniGenerationScheduler
-      gpu_memory_utilization: 0.15
-      enforce_eager: true
-      trust_remote_code: true
-      enable_prefix_caching: false
-      engine_output_type: audio
-    engine_input_source: [1]
-    final_output: true
-    final_output_type: audio
-    default_sampling_params:
-      temperature: 0.0
-      top_p: 1.0
-      top_k: -1
-      max_tokens: 2048
-      seed: 42
-      detokenize: True
-      repetition_penalty: 1.1
-
-# Top-level runtime config (concise): default windows and stage edges
-runtime:
-  enabled: true
-
-  edges:
-    - from: 0                   # thinker → talker: trigger only after receiving full input (-1)
-      to: 1
-    - from: 1                   # talker → code2wav: trigger only after receiving full input (-1)
-      to: 2
-
-```
-
-## Legacy stage configuration arguments
-
-Each stage in the `stage_args` list contains the following configuration options:
-
-### `stage_id`
-
-A unique identifier for each stage in the multi-stage pipeline. Stages are numbered sequentially starting from 0, and this ID is used to reference stages in inter-stage dependencies (e.g., `engine_input_source`).
-
-### `prompt_expand_func` (Optional)
-
-A custom Python function hook for the LLM stage (Stage 0) that expands a single incoming prompt object into multiple prompts. This is primarily used for multi-modal Classifier-Free Guidance (CFG), where it generates the necessary companion requests (like a negative text prompt) and tags them with internal roles (e.g., `cfg_text`). This ensures the upstream LLM generates the needed contextual hidden states for both the conditional and unconditional generations simultaneously.
-
-### `cfg_kv_collect_func` (Optional)
-
-A custom Python function hook for downstream diffusion stages (Stage 1+) to collect, map, and process the KV caches transferred from the companion requests fired by `prompt_expand_func`. It aggregates the hidden condition states cleanly (e.g., binding them as `cfg_text_past_key_values` and `cfg_text_kv_metadata`), allowing the diffusion runtime to perform CFG smoothly without redundantly evaluating text paths on the DiT workers.
-
-### `runtime`
-
-Configuration for disaggregated execution of the stage, controlling how the stage is deployed and executed.
-
-#### `runtime.process`
-
-Whether to run this stage in a separate process. When set to `true`, the stage will be executed in an isolated process, enabling better resource isolation and parallel execution across different stages. This is essential for multi-GPU deployments where different stages run on different devices.
-
-Default: `true`
-
-#### `runtime.devices`
-
-Logical device indices for this stage, specified as a string. Values are **logical indices** (`0`, `1`, `2`, ...) — not physical GPU IDs — and are mapped through the platform's visibility env var (`CUDA_VISIBLE_DEVICES` on CUDA, `ASCEND_RT_VISIBLE_DEVICES` on NPU) before being applied via `torch.cuda.set_device()` (or the equivalent).
-
-Example: if `CUDA_VISIBLE_DEVICES=0,2,4` is set in the environment, then `devices: "0"` selects physical GPU 0 (the first visible), `devices: "1"` selects physical GPU 2, and `devices: "0,1"` makes physical GPUs 0 and 2 available to the stage. If no visibility env var is set, logical and physical IDs coincide.
-
-Default: `"0"`
-
-#### `engine_args.max_num_seqs`
-
-The maximum number of sequences for concurrent processing in this stage. For
-LLM stages, this controls the vLLM scheduler's maximum concurrent sequences.
-For diffusion stages, this controls scheduler wave capacity in both request
-batch mode and step batch mode. `max_num_seqs: 1` is the serial/conservative
-path; values above `1` allow compatible requests to batch when the selected
-diffusion pipeline supports that batching mode.
-
-Default: `1`
-
-#### `engine_args.request_batch_max_wait_ms`
-
-The maximum time, in milliseconds, that a diffusion request-mode stage may wait
-before the first `schedule()` of a new scheduler wave so compatible requests can
-accumulate for request-level batching. This only applies to diffusion pipelines
-that support request-level batching with `step_execution` disabled.
-
-Use this together with `max_num_seqs > 1` for bursty serving traffic. `0`
-disables admission waiting and preserves the lowest first-request latency.
-For diffusion execution and batching tuning, see
-[Diffusion Execution Modes](../user_guide/diffusion/execution_modes.md).
-
-Default: `0.0`
-
-### `engine_args`
-
-Engine arguments for configuring the LLM engine, diffusion engine, or other engine types used by this stage.
-
-#### `engine_args.model_stage`
-
-The name identifier for this model stage within the multi-stage architecture. This is used internally to distinguish different stages of the same model (e.g., "thinker", "talker", "code2wav" in Qwen2.5-Omni).
-
-#### `engine_args.model_arch`
-
-The model architecture class name that is registered in `model_executor/models/registry.py`. This specifies which model implementation to use for this stage. The class must be registered in the model registry for vLLM-Omni to locate and instantiate it.
-
-#### `engine_args.worker_cls`
-
-The specific worker class to use for this stage. This determines how the model computations are executed. Examples include `vllm_omni.worker.gpu_ar_worker.GPUARWorker` for autoregressive stages and `vllm_omni.worker.gpu_generation_worker.GPUGenerationWorker` for diffusion-based stages.
-
-#### `engine_args.scheduler_cls`
-
-The scheduler class to use for this stage. The scheduler manages request queuing, batching, and execution order. Examples include `vllm_omni.core.sched.omni_ar_scheduler.OmniARScheduler` for standard stages and `vllm_omni.core.sched.omni_generation_scheduler.OmniGenerationScheduler` for diffusion stages.
-
-#### `engine_args.gpu_memory_utilization`
-
-The fraction of GPU memory to allocate for this stage within a single GPU chip. This is a value between 0.0 and 1.0, where 0.8 means 80% of the GPU memory will be used by this stage. This allows fine-grained control over memory allocation when multiple stages share the same GPU or when reserving memory for other operations.
-
-Default: `0.8`
-
-!!! tip "Memory Configuration Guide"
-    For detailed information on how to calculate memory requirements and properly configure `gpu_memory_utilization`, see the [GPU Memory Calculation and Configuration Guide](./gpu_memory_utilization.md).
-
-#### `engine_args.enforce_eager`
-
-Whether to enforce eager execution mode. When set to `true`, the engine will run in eager mode without using CUDA graphs or other compilation optimizations. Currently, vLLM-Omni only supports eager mode.
-
-Default: `true`
-
-#### `engine_args.trust_remote_code`
-
-Whether to trust remote code when loading models from Hugging Face. This is required for models that use custom code in their configuration files. Set to `true` when loading models that require custom model implementations.
-
-Default: `true`
-
-#### `engine_args.engine_output_type`
-
-Specifies the type of output produced by this stage's engine. This determines what kind of data flows to downstream stages. Possible values include `latent` (hidden states), `text` (tokenized text), and `audio` (audio waveforms). When set to `latent`, the stage outputs latent hidden states in addition to token IDs, which are consumed by downstream stages.
-
-Default: `latent`
-
-#### `engine_args.enable_prefix_caching`
-
-Whether to enable prefix caching for this stage. Prefix caching can improve performance by caching KV cache for common prompt prefixes. However, for requests that output hidden states (when `engine_output_type` is `latent`), prefix caching is not currently supported and should be set to `false`.
-
-Default: `false`
-
-### `is_comprehension`
-
-Whether this stage is a text or multimodal comprehension module. When set to `true`, the stage acts as a comprehension module that processes input text or multimodal content. If this is the first comprehension stage, `AsyncOmni` will use its tokenizer as the default tokenizer for the entire pipeline.
-
-Default: `true`
-
-### `final_output`
-
-Whether this stage produces output that is part of the final outputs returned to the user. When set to `false`, the stage only works as an intermediate stage, processing data that flows to downstream stages but not contributing directly to the final response.
-
-Default: `true`
-
-### `final_output_type`
-
-The type of final output produced by this stage. This specifies what format the output will be in when returned to the user. Currently supported values are `text` (for text generation) and `audio` (for audio generation).
-
-Default: `text`
-
-### `default_sampling_params`
-
-Default sampling parameters for this stage. These parameters control the generation behavior and align with vLLM's sampling parameter semantics. These defaults are used when no explicit sampling parameters are provided in the request.
-
-#### `default_sampling_params.temperature`
-
-Sampling temperature for controlling randomness. Lower values (e.g., 0.0) make the output more deterministic and focused, while higher values increase randomness.
-
-Default: `0.0`
-
-#### `default_sampling_params.top_p`
-
-Nucleus sampling parameter. Only tokens with cumulative probability mass up to `top_p` are considered. This helps filter out low-probability tokens while maintaining diversity.
-
-Default: `1.0`
-
-#### `default_sampling_params.top_k`
-
-Top-k sampling parameter. Only the top `k` most likely tokens are considered. Set to `-1` to disable top-k filtering and consider all tokens.
-
-Default: `-1`
-
-#### `default_sampling_params.max_tokens`
-
-Maximum number of tokens to generate in this stage. This limits the length of the output sequence.
-
-Default: `2048`
-
-#### `default_sampling_params.seed`
-
-Random seed for reproducible generation. When set, the random number generator will be initialized with this seed to ensure consistent outputs across runs.
-
-Default: `42`
-
-#### `default_sampling_params.detokenize`
-
-Whether to detokenize the output tokens into text. When set to `true`, token IDs are converted back to readable text strings.
-
-Default: `True`
-
-#### `default_sampling_params.repetition_penalty`
-
-Penalty applied to tokens that have already appeared in the generated sequence. Values greater than 1.0 discourage repetition, while values less than 1.0 encourage it. A value of 1.0 applies no penalty.
-
-Default: `1.1`
-
-### `tts_args` (TTS stages only)
-
-Configuration for Text-to-Speech specific parameters. This section is only applicable to TTS model stages (e.g., `qwen3_tts`).
-
-#### `tts_args.max_instructions_length`
-
-Maximum character length for voice style/emotion instructions. Instructions exceeding this limit will be rejected with a validation error.
-
-Default: `500`
-
-This value can be overridden at runtime using the `--tts-max-instructions-length` CLI parameter when starting the server.

docs/contributing/DOCS_GUIDE.md

@@ -46,7 +46,7 @@ class Omni:
 
     Args:
         model: Model name or path
-        stage_configs: Optional stage configurations
+        deploy_config: Optional path to a deploy configuration
         **kwargs: Additional arguments passed to the engine
 
     Example:

docs/contributing/README.md

@@ -44,6 +44,149 @@ vLLM-Omni's `pre-commit` hooks will now run automatically every time you commit.
     pre-commit run --show-diff-on-failure --color=always --all-files  # runs on all files (short for --all-files)
     ```
 
+!!! warning
+    GitHub Actions `pre-commit` **skips** several local gates so a whole-tree
+    `--all-files` run stays green while historical debt is cleaned up. The
+    current `SKIP` list in
+    [`.github/workflows/pre-commit.yml`](https://github.com/vllm-project/vllm-omni/blob/main/.github/workflows/pre-commit.yml)
+    is `check-test-ci-coverage`, `markdownlint-cli2`, `shellcheck`,
+    `check-spdx-header`, and `mypy-3.10`. Those hooks still run on **your
+    commit** for changed files. A passing GitHub pre-commit check does not mean
+    they passed locally. Hooks **not** on that list (forbidden imports,
+    `torch.cuda`, TTS adapter ratchet, Buildkite schema, Ruff, typos, …) run in
+    both places.
+
+The hooks below are new relative to the previous Omni config (Ruff, typos,
+actionlint, YAML/whitespace, DCO sign-off, and the pickle-only checker were
+already there). `check-pickle-imports` is gone: pickle is now one rule inside
+`check-forbidden-imports`.
+
+| Hook id | Enforces | GitHub Actions |
+| ------- | -------- | -------------- |
+| `markdownlint-cli2` | Markdown in `docs/`, `recipes/`, `README.md`, `CONTRIBUTING.md` (not `.claude/` / `.cursor/` / `CLAUDE.md`) | skipped |
+| `mypy-3.10` | Type-check changed `vllm_omni/` files (model trees excluded). `tests/` uses `--follow-imports skip` | skipped |
+| `mypy-3.11` / `3.12` / `3.13` | Same checker, extra Python versions | not installed; `pre-commit run --hook-stage manual mypy-3.12` |
+| `check-test-ci-coverage` | Every `tests/**/test_*.py` has a CI level mark and a hardware mark/helper | skipped |
+| `check-tts-adapter-migration` | `self._tts_model_type` branches in `serving_speech.py` must not increase | runs |
+| `shellcheck` | `*.sh` quoting / undefined vars | skipped |
+| `check-spdx-header` | Omni SPDX header on `.py` / `.pyi` / `.sh` / `.rs` / `.proto` | skipped |
+| `check-forbidden-imports` | pickle, stdlib `re`/`base64`, Triton, TileLang, Hugging Face Hub download/API | runs |
+| `check-torch-cuda-call` | Direct `torch.cuda.*` helpers outside platform adapters | runs |
+| `check-buildkite` | `.buildkite/*.yml` against the Buildkite schema | runs |
+
+Related config (not a new hook id): Ruff `TID251` also bans `librosa` and several
+`torch.cuda.*` names; typos is pinned and ignores git-describe hashes (`+g` +
+hex) and NumPy `writeable`. Keep the `suggestion` hook last in
+`.pre-commit-config.yaml` so autofix output does not bury the SKIP tip.
+
+#### SPDX headers
+
+Source files must start with:
+
+```text
+# SPDX-License-Identifier: Apache-2.0
+# SPDX-FileCopyrightText: Copyright contributors to the vLLM-Omni project
+```
+
+Rust and proto files use `//` instead of `#`. Empty files are exempt. A stale
+`Copyright contributors to the vLLM project` line is **rewritten in place** to
+the Omni copyright; the hook then exits non-zero so you restage the rewrite.
+
+#### Forbidden imports
+
+Under `vllm_omni/`, do not use:
+
+- stdlib `re` → `import regex as re`
+- stdlib `base64` → `import pybase64` (or `import pybase64 as base64`)
+- `pickle` / `cloudpickle` (also banned in most tests)
+- `from huggingface_hub import snapshot_download` / `hf_hub_download` / `HfApi`
+  / similar → `vllm.transformers_utils.repo_utils`
+- direct `triton` / `tilelang` → `vllm.triton_utils` / `vllm.tilelang_utils`
+
+`examples/`, `tests/`, `benchmarks/`, `apps/`, `docs/`, `tools/`, `scripts/`,
+`.buildkite/`, and `.github/` may keep stdlib `re` and `base64`
+(`_NON_LIBRARY_DIRS`). Pickle is stricter: tests are not exempt unless the
+file is on the pickle allowlist.
+
+#### `torch.cuda` call sites
+
+Prefer `current_omni_platform` / `OmniPlatform`. Platform adapters under
+`vllm_omni/platforms/` are allowed; other files must not add new
+`torch.cuda.*` helpers (`device_count`, `synchronize`, `empty_cache`,
+`device()`, `manual_seed`, and similar). `import torch` is fine.
+
+#### shellcheck
+
+The hook does **not** download a binary. Install `shellcheck` with a signed
+package manager, then re-run:
+
+- Debian/Ubuntu/WSL: `sudo apt-get install shellcheck`
+- Fedora: `sudo dnf install ShellCheck`
+- macOS: `brew install shellcheck`
+- Git Bash (MINGW): `scoop install shellcheck` so `shellcheck.exe` is on PATH.
+  WSL is Linux: use `apt-get`, not a Windows `.exe`.
+
+If no native binary is found, the wrapper prints those commands (and
+[the upstream install page](https://github.com/koalaman/shellcheck?tab=readme-ov-file#installing))
+and exits 1. AMD and other vendor scripts are **not** excluded.
+
+#### mypy
+
+`mypy-3.10` runs on commit for changed files. `[tool.mypy]` uses Python 3.12,
+`ignore_missing_imports = true`, and `follow_imports = "silent"`. Wrapper
+`tools/pre_commit/mypy.py` skips `vllm_omni/model_executor/models/` and
+`vllm_omni/diffusion/models/`, and type-checks `tests/` with
+`--follow-imports skip`. Extra versions:
+
+```bash
+pre-commit run --hook-stage manual mypy-3.12
+```
+
+#### Test CI marks
+
+`check-test-ci-coverage` requires each collected test module under `tests/` to
+have at least one CI **level** mark (`core_model`, `advanced_model`,
+`full_model`, `local_model`, or `slow`) and a **hardware** mark (`cpu`, `cuda`,
+`H100`, …) or helper (`hardware_test(` / `hardware_marks(`). GitHub Actions
+skips this hook; local commit does not. See the
+[test writing guide](./ci/test_writing_guide.md).
+
+#### TTS adapter ratchet
+
+`check-tts-adapter-migration` counts `self._tts_model_type` comparisons in
+`vllm_omni/entrypoints/openai/serving_speech.py`. The count must stay at or
+below `MAX_MODEL_TYPE_BRANCHES` in
+[`tools/pre_commit/check_tts_adapter.py`](https://github.com/vllm-project/vllm-omni/blob/main/tools/pre_commit/check_tts_adapter.py).
+New per-model behavior belongs in
+`vllm_omni/entrypoints/openai/tts_adapters/`. Removing branches **must** lower
+the budget in the same change; raising it is a reviewable policy edit.
+
+#### Markdownlint and Buildkite
+
+`markdownlint-cli2` auto-fixes `docs/`, `recipes/`, and the root README /
+CONTRIBUTING (rules in `.markdownlint.yaml`). Skill files under `.claude/` and
+`.cursor/` are excluded.
+
+`check-buildkite` validates `.buildkite/*.yml` against the official schema
+after the same expansion `upload_pipeline.py` uses. Omni-only keys such as
+`mirror_hardwares` are stripped or rendered first. A few files are skipped in
+`SKIP_FILES` inside `tools/pre_commit/check_buildkite.py`.
+
+#### Extending allowlists and budgets
+
+Prefer fixing the call site. Growing an allowlist or raising a budget is a
+policy change and needs review — do not edit these just to make the hook pass.
+
+| Gate | File | What to edit |
+| ---- | ---- | ------------ |
+| pickle, `re`, `base64`, Hugging Face Hub, Triton/TileLang | [`tools/pre_commit/check_forbidden_imports.py`](https://github.com/vllm-project/vllm-omni/blob/main/tools/pre_commit/check_forbidden_imports.py) | `CHECK_IMPORTS[<rule>].allowed_files`. Use `allowed_dirs` only when a whole tree should stay exempt. |
+| `torch.cuda` | [`tools/pre_commit/check_torch_cuda.py`](https://github.com/vllm-project/vllm-omni/blob/main/tools/pre_commit/check_torch_cuda.py) | `ALLOWED_FILES`, or `ALLOWED_PREFIXES` for a platform adapter tree. |
+| TTS `_tts_model_type` branches | [`tools/pre_commit/check_tts_adapter.py`](https://github.com/vllm-project/vllm-omni/blob/main/tools/pre_commit/check_tts_adapter.py) | `MAX_MODEL_TYPE_BRANCHES` (down only, unless the PR justifies a raise). |
+| Buildkite skip list | [`tools/pre_commit/check_buildkite.py`](https://github.com/vllm-project/vllm-omni/blob/main/tools/pre_commit/check_buildkite.py) | `SKIP_FILES` for pipelines that cannot be expanded. |
+
+To skip one hook for a single commit (discouraged): `SKIP=<hook-id> git commit`.
+To bypass every hook: `git commit --no-verify` (also discouraged).
+
 ### Documentation
 
 MkDocs is a fast, simple and downright gorgeous static site generator that's geared towards building project documentation. Documentation source files are written in Markdown, and configured with a single YAML configuration file, `mkdocs.yml`.
@@ -162,12 +305,14 @@ Only specific types of PRs will be reviewed. The PR title is prefixed appropriat
 Before submitting a PR, run the [precheck-pr skill](https://github.com/vllm-project/vllm-omni/blob/main/.claude/skills/precheck-pr/SKILL.md) with the code agent for a self-review against project conventions:
 
 The skill offers two modes:
+
 - **Quick (~3 min):** catches showstoppers — PR title format, missing benchmark claims, rebase status
 - **Full (~10 min):** thorough maintainer-grade review — dead code scan, copy-paste detection, import hygiene
 
 The precheck covers five PR types: Bug Fix, Performance, New Model, Diffusion Model, and General. Each type has a tailored checklist that validates evidence quality (repro steps, A/B benchmarks, registry entries, etc.). See the [precheck-pr skill](https://github.com/vllm-project/vllm-omni/blob/main/.claude/skills/precheck-pr/SKILL.md) for the full checklist.
 
 ### Local Test
+
 Please run the L1 and L2 test cases locally first and attach the results before contacting us to add the "ready" label. Please refer to the [test instructions](./ci/test_execution_guide.md) for running the test cases.
 
 ### Automatic skip-ci (docs and pytest skip marks)
@@ -189,7 +334,7 @@ These PR shapes all trigger skip-ci:
 Skip-ci does **not** apply when the diff also touches product code (for example `vllm_omni/`), or when test files change assertions, imports, fixtures, or other non-skip logic. If the diff cannot be resolved (non-PR branches outside `main`), CI runs as usual.
 
 !!! note
-    Skipping L2/L3 does **not** disable the Docker image build step. Nightly (L4) upload can still run when the PR has a `nightly-test` label or on scheduled `main` builds with `NIGHTLY=1`. Bootstrap child steps live in `bootstrap-upload-steps.yml`; the hook entry step runs `upload_pipeline.py --upload <platform>/bootstrap-upload-steps.yml`, which injects `if` by step `key` from skip-ci before upload. See [CI Settings — Diff-aware CI](./ci/ci_settings.md#diff-aware-ci) and [Test System Overview](./ci/test_system_overview.md).
+    Skipping L2/L3 does **not** disable the Docker image build step. Nightly (L4) upload can still run when the PR has a `nightly-test` label or on scheduled `main` builds with `NIGHTLY=1`. Weekly (L5) upload can still run with `weekly-test`, `WEEKLY=1`, or `NON_CRITICAL=1`. Bootstrap child steps live in `bootstrap-upload-steps.yml`; the hook entry step runs `upload_pipeline.py --upload <platform>/bootstrap-upload-steps.yml`, which injects `if` by step `key` from skip-ci before upload. See [CI Settings — Diff-aware CI](./ci/ci_settings.md#diff-aware-ci) and [Test System Overview](./ci/test_system_overview.md).
 
 ### Code Quality
 

docs/contributing/ci/ci_settings.md

@@ -8,7 +8,7 @@ For CI levels (L1–L5) and triggers, see [Test System Overview](./test_system_o
 
 Canonical layout (prefer these paths for new changes):
 
-```
+```text
 .buildkite/
 ├── common/                          # Shared across platforms
 │   ├── scripts/
@@ -48,7 +48,7 @@ Canonical layout (prefer these paths for new changes):
     └── scripts/
 ```
 
-**Placement rules**
+### Placement rules
 
 | Rule | Detail |
 | ---- | ------ |
@@ -81,7 +81,14 @@ There are still **legacy copies** at `.buildkite/*.yaml` (without the `cuda/` pr
     | L2 | `cuda/test-ready.yml` | `ready` label |
     | L3 | `cuda/test-merge.yml` | `merge-test` label / main merge |
     | L4 | `cuda/test-nightly.yml` | `nightly-test` label or `NIGHTLY=1` |
-    | L5 | `cuda/test-weekly.yml` | `weekly-test` label or `WEEKLY=1` |
+    | L5 | `cuda/test-weekly.yml` | `weekly-test` label, `WEEKLY=1`, or `NON_CRITICAL=1` (see step `if`s below) |
+
+    On scheduled `main` builds, `test-weekly.yml` is uploaded when **`WEEKLY=1` or `NON_CRITICAL=1`**. Inside that file:
+
+    - **Reliability** / **Perf Test** / **Simple · CPU Coverage Test** → `WEEKLY=1` (or PR label `weekly-test` for Reliability/Perf). Pipeline upload for scheduled env vars still requires `main`.
+    - **E2E Tests** group (slow Omni/TTS/Diffusion sweeps) → `NON_CRITICAL=1`
+
+    **`--e2e`:** when uploading ready/merge on `main` with **`WEEKLY=1`**, bootstrap passes `--e2e` so only the **E2E Test** group is kept (same flag `run_cov_split.sh` uses to enable per-model coverage).
 
     **Upload:** `upload_pipeline.py --upload` expands uploader-only keys before Buildkite upload.
 
@@ -112,6 +119,8 @@ There are still **legacy copies** at `.buildkite/*.yaml` (without the `cuda/` pr
 
     **Test YAML:** `npu/test-npu-ready.yml` (L2), `test-npu-nightly.yml` (L4).
 
+    **L2 trigger:** PR label `ready` (CUDA/AMD L2). PR labels `ready` + `npu-test` trigger NPU CI only.
+
     **Upload:** same as CUDA—`upload_pipeline.py --upload`.
 
     **Hardware in YAML:** `mirror_hardwares` preset (string), expanded to `agents`, top-level `image`, and `plugins`. Presets: `a2b3_npu_1`, `a2b3_npu_4`, `a3_npu_2` in `common/ci_mirror_hardwares.yml`.
@@ -204,15 +213,15 @@ Label triggers (`ready`, `merge-test`) are unchanged—diff-aware logic only red
 | Diff shape | Path | Default L2/L3 |
 | --- | --- | --- |
 | Product code / non-whitelisted paths | normal CI | all on |
-| Docs and/or qualifying skip-mark only | `skip_all` | all off (scheduled `main` NIGHTLY/WEEKLY exceptions below) |
+| Docs and/or qualifying skip-mark only | `skip_all` | all off (scheduled `main` NIGHTLY/WEEKLY/`NON_CRITICAL` exceptions below) |
 | Whitelisted CI YAML only | yaml-gated (`skip_l2_l3`) | per platform/level matrix |
 | Docs/skip-mark **+** whitelisted CI YAML | yaml-gated (`skip_l2_l3`) | same as CI-YAML-only (does **not** widen to normal CI) |
 | Non-qualifying skip-mark and no CI YAML | normal CI | all on |
 | Diff unavailable | normal CI | all on |
 
-**`skip_all` exceptions (CUDA/NPU bootstrap only):** PR labels (`nightly-test`, `merge-test`, `npu-test`, `weekly-test`, …) do **not** revive jobs. On `main`, scheduled `NIGHTLY=1` still builds the image and uploads L4 plus L2/L3 (with `--e2e`); `WEEKLY=1` still builds the image and uploads L5 (CUDA).
+**`skip_all` exceptions (CUDA/NPU bootstrap only):** PR labels (`nightly-test`, `merge-test`, `weekly-test`, `ready`, …) do **not** revive jobs under `skip_all`. On `main`, scheduled `NIGHTLY=1` still builds the image and uploads L4; `WEEKLY=1` or `NON_CRITICAL=1` still builds the image and uploads L5 (CUDA). `WEEKLY=1` on `main` also uploads L2/L3 with `--e2e`.
 
-**Yaml-gated nightly/weekly:** L4/L5 upload steps keep their normal label / `NIGHTLY` / `WEEKLY` conditions (for example PR `nightly-test` still uploads nightly). Only L2/L3 upload steps are matrix-gated.
+**Yaml-gated nightly/weekly:** L4/L5 upload steps keep their normal label / `NIGHTLY` / `WEEKLY` / `NON_CRITICAL` conditions (for example PR `nightly-test` still uploads nightly). Only L2/L3 upload steps are matrix-gated.
 
 ##### Whitelisted CI YAML → platform
 
@@ -285,12 +294,12 @@ CUDA **L2** (`.buildkite/cuda/test-ready.yml`) and **L3** (`.buildkite/cuda/test
 | `source_file_dependencies` | List of path **prefixes**. If any changed file equals a prefix or starts with `prefix/`, keep the step (or group); otherwise omit it. |
 | `mirror_hardwares` | Expand to `agents` + `plugins` (+ optional `image`) from `ci_mirror_hardwares.yml`. |
 
-**Policy**
+### Policy
 
 - **Always uploaded** (no key): groups outside **E2E Test**—Simple Test, Diffusion unit tests, Engine/Model Executor, Distributed, Custom Pipeline, Entrypoints (L2), LoRA / Entrypoints (L3).
 - **Diff-gated**: every **E2E Test** leaf job. List the smallest prefix set per step—pytest file(s), model code under `vllm_omni/model_executor/models/` or `vllm_omni/diffusion/models/`, plus `stage_input_processors/` and `vllm_omni/deploy/*.yaml` when applicable. A **group** may define the key instead; the whole group drops if no prefix matches.
 
-**YAML examples**
+### YAML examples
 
 ```yaml
       - label: "Diffusion · Qwen Image Test"
@@ -312,7 +321,7 @@ CUDA **L2** (`.buildkite/cuda/test-ready.yml`) and **L3** (`.buildkite/cuda/test
         mirror_hardwares: l4_4
 ```
 
-**Local dry-run**
+### Local dry-run
 
 ```bash
 python3 .buildkite/common/scripts/upload_pipeline.py .buildkite/cuda/test-ready.yml
@@ -324,16 +333,19 @@ On PR builds, `upload_pipeline.py` logs `skip '…' (no changes under …)` for
 
 ### Labels and grouping
 
-- Use **groups** for dashboard readability; keep **E2E Test** as the group name CUDA diff filtering expects for `--e2e` nightly runs on main.
+- Use **groups** for dashboard readability; keep **E2E Test** as the group name CUDA diff filtering expects for `--e2e` weekly runs on main (`WEEKLY=1`).
 - Prefix labels by model domain: **Omni ·**, **TTS ·**, **Diffusion ·**, **Simple ·**, etc., matching existing steps.
 
 ### Per-model coverage
 
-Pilot (v1): two Merge-tier (L3) jobs in `.buildkite/cuda/test-merge.yml` upload
-per-model, per-entry-mode coverage as Buildkite artifacts — "Diffusion · Bagel
-Test" and "TTS · Qwen3-TTS Base Test". Both run on a single GPU so the pilot is
-cheap to reproduce. This is a pilot, not a rollout: no dashboard/visualization
-lives in this repo, and nothing is gated on coverage.
+Pilot (v1): one Merge-tier (L3) job in `.buildkite/cuda/test-merge.yml` uploads
+per-model, per-entry-mode coverage as Buildkite artifacts — "TTS · Qwen3-TTS Base
+Test". Coverage + artifact upload run only when bootstrap uploaded that job with
+`--e2e` (`WEEKLY=1` and `BUILDKITE_BRANCH=main`, via `run_cov_split.sh`); otherwise
+the same job runs plain pytest without `--cov`. It runs on a single GPU so the
+pilot is cheap to reproduce.
+
+**L1 package coverage (CPU):** ready/merge Simple Test jobs run without `--cov`. Scheduled weekly on `main` with `WEEKLY=1` runs **Simple · CPU Coverage Test** in `.buildkite/cuda/test-weekly.yml` (`pytest -m 'core_model and cpu'` + Cobertura XML artifact).
 
 **Naming convention**: `coverage-<model_id>-<mode>-<step_id>.xml.gz`, where
 `<model_id>` is the model's directory name under
@@ -352,8 +364,8 @@ imported. A run covering zero lines produces the same 7 MB. gzip takes it to
 roughly 410 KB and changes nothing about the data.
 
 `<step_id>` is what keeps two steps that cover the same model apart. Artifacts are
-scoped to the build, and a nightly build carries the ready, merge and nightly tiers
-at once, so several same-model steps land in one namespace — `qwen3_tts` already has
+scoped to the build, and a weekly build can carry ready/merge (with `--e2e`) and
+weekly tiers at once, so several same-model steps land in one namespace — `qwen3_tts` already has
 both a Base and a CustomVoice job, and `minicpmo_4_5` a base and a duplex one. Two
 uploads on one path do not fail loudly: an exact-name download can report the path
 as ambiguous, and a glob download fetches every match concurrently and renames each
@@ -410,7 +422,8 @@ runs a coverage job needs the same.
 List both `run_cov_split.sh` and `pyproject.toml` in every opted-in job's
 `source_file_dependencies` — both change what the job measures, so without them a
 change there is filtered out of normal PR builds and only surfaces in a later
-nightly. `tests/buildkite/test_upload_pipeline.py` asserts this for the pilots.
+nightly. `tests/buildkite/test_upload_pipeline.py` covers the filter behavior with
+a synthetic job (it does not pin real merge labels).
 Editing only the surrounding CI YAML still does not schedule them, so a PR that
 touches just the wiring needs a full E2E run (or the commands run on a GPU host)
 to produce artifacts. When checking a new model's
@@ -422,7 +435,7 @@ just confirming both files exist.
 | Check | Command / location |
 | ----- | ------------------ |
 | CUDA render | `python3 .buildkite/common/scripts/upload_pipeline.py .buildkite/cuda/test-<level>.yml` |
-| No leaked uploader keys | `… \| grep -E 'mirror_hardwares|source_file_dependencies'` → empty |
+| No leaked uploader keys | Pipe the render into `grep -E 'mirror_hardwares\|source_file_dependencies'` — expect empty |
 | Skip-ci / upload unit tests | `pytest tests/buildkite/` |
 | Local job replay (CUDA L2+) | `tools/run_ready_jobs.sh`, `tools/run_merge_jobs.sh`, `tools/nightly/run_nightly_jobs.sh` (read YAML from `cuda/`) |
 

docs/contributing/ci/failures.md

@@ -6,7 +6,7 @@ When you open a PR against vLLM-Omni, several CI checks run automatically:
 
 | Check | Platform | What it does |
 | ----- | -------- | ------------ |
-| **pre-commit** | GitHub Actions | Runs linting (Ruff), formatting, spell-checking (typos), and YAML validation. |
+| **pre-commit** | GitHub Actions | Runs Ruff, formatting, typos, YAML validation, forbidden imports, `torch.cuda`, TTS adapter ratchet, and Buildkite schema. Local-only gates (SPDX, shellcheck, markdownlint, mypy-3.10, test marks) are in the workflow `SKIP` env — see [Linting](../README.md#linting). |
 | **Build Wheel** | GitHub Actions | Builds Python wheels for Python 3.11 and 3.12 on Ubuntu. Skipped for docs-only or Markdown-only changes (controlled by `paths-ignore` in the workflow). |
 | **DCO** | GitHub | Verifies every commit has a `Signed-off-by` line. |
 | **docs/readthedocs.org:vllm-omni** | Read the Docs | Builds the MkDocs documentation site. |
@@ -29,6 +29,13 @@ pre-commit run --all-files
 
 Then commit the fixes and push.
 
+A green GitHub pre-commit job can still miss SPDX, shellcheck, mypy-3.10,
+markdownlint, and test-mark coverage because CI skips those IDs. If your
+**local** commit hook fails, follow [Linting](../README.md#linting) for the
+full hook list (forbidden imports, `torch.cuda`, TTS adapter ratchet, and
+Buildkite schema still run in GHA). Do not grow allowlists or raise
+`MAX_MODEL_TYPE_BRANCHES` without review.
+
 ### DCO failures
 
 Every commit must include a `Signed-off-by` line. If you forgot, amend your commits:

docs/contributing/ci/test_examples/l4_functionality_tests.inc.md

@@ -1,8 +1,8 @@
-**Scope**
+#### Scope
 
 For diffusion models, the L4 functionality test covers all or common *diffusion features* that are supported by this model, including several [parallelism acceleration methods](https://docs.vllm.ai/projects/vllm-omni/en/latest/user_guide/diffusion/parallelism_acceleration/), [CPU offloading](https://docs.vllm.ai/projects/vllm-omni/en/latest/user_guide/diffusion/cpu_offload/), [TeaCache](https://docs.vllm.ai/projects/vllm-omni/en/latest/user_guide/diffusion/teacache/) and [Cache-DiT](https://docs.vllm.ai/projects/vllm-omni/en/latest/user_guide/diffusion/cache_dit_acceleration/) cache backends, [quantization methods](https://docs.vllm.ai/projects/vllm-omni/en/latest/user_guide/quantization/overview/).
 
-**Test Case Design**
+#### Test Case Design
 
 For a *high priority* model (currently listed in [issue #1832](https://github.com/vllm-project/vllm-omni/issues/1832)), we use several test cases, each with multiple features turned on, so that each supported feature is tested in at least one test case. This is to relieve the GPU workload on the CI machine. The suggested test case combination is as follows:
 
@@ -34,12 +34,12 @@ For a *normal priority* model, further reduce the number of test cases.
 
 Currently all the features are available in online serving mode. Hence, only need to add `tests/e2e/online_serving/test_{model}_expansion.py`.
 
-**Code Style**
+#### Code Style
 
-- Validation: test that the multimodal output files of your model have the correct shapes. `OpenAIClientHandler.send_diffusion_request` should have taken care of this.
+- Validation: test that the multimodal output files of your model have the correct shapes. `OnlineOmniClient.send_diffusion_request` should have taken care of this.
 - Test marks: always add `full_model` and `diffusion` for L4 nightly `test_*_expansion.py` cases. Add GPU-related marks if needed. Ref: [Markers for Tests](https://docs.vllm.ai/projects/vllm-omni/en/latest/contributing/ci/test_writing_guide/#markers-for-tests).
 - To maximize code reuse, you may refer to
-    - `tests/conftest.py` for `omni_server` (running server in subprocess) and `openai_client` fixtures (sending requests and validating output), `generate_synthetic_image` and `assert_XXX_valid` helper.
+    - `tests/conftest.py` for `omni_server` (running server in subprocess) and `online_client` fixtures (sending requests and validating output), `generate_synthetic_image` and `assert_XXX_valid` helper.
     - `tests/helpers/mark.py` for `@hardware_test(...)` and `hardware_marks`.
     - [Parametrizing tests (pytest doc)](https://docs.pytest.org/en/stable/example/parametrize.html) to reuse test function implementation for different cases.
 - Doc: add a concise docstring for each test function.

docs/contributing/ci/test_examples/l4_performance_tests.inc.md

@@ -1,14 +1,15 @@
 When you want to add L4-level ***performance test*** cases, add entries to JSON files under `tests/dfx/perf/tests/` and run them via `tests/dfx/perf/scripts/run_benchmark.py` (omni / TTS) or `run_diffusion_benchmark.py` (diffusion).
 
-**Config file layout (in-tree examples)**
+#### Config file layout (in-tree examples)
 
 | Model type | Runner | Example JSON files |
 | ---------- | ------ | ------------------ |
-| Omni | `run_benchmark.py` | `test_qwen3_omni_no_async_chunk.json`, `test_qwen3_omni_async_chunk.json`, `test_qwen3_omni_vllm_text.json`, `test_qwen3_omni_multi_replicas.json` |
+| Omni (nightly) | `run_benchmark.py` | `test_qwen3_omni_no_async_chunk.json`, `test_qwen3_omni_async_chunk.json` (`full_model` in `mark`) |
+| Omni (weekly) | `run_benchmark.py` | `test_qwen3_omni_vllm_text.json`, `test_qwen3_omni_multi_replicas.json` (`slow` in `mark`; **Perf Test** in `test-weekly.yml`) |
 | TTS | `run_benchmark.py` | `test_tts.json`, `test_voxcpm2.json`, `test_higgs_audio_v3.json` |
 | Diffusion | `run_diffusion_benchmark.py` | `test_qwen_image_vllm_omni.json`, `test_bagel_vllm_omni.json`, `test_wan22_i2v_vllm_omni.json`, `test_cosmos3_vllm_omni.json`, … |
 
-**How runners pick cases**
+#### How runners pick cases
 
 Without **`--test-config-file`**, each runner scans all `*.json` under `tests/dfx/perf/tests/` but only keeps its own model type:
 
@@ -17,9 +18,9 @@ Without **`--test-config-file`**, each runner scans all `*.json` under `tests/df
 
 Diffusion cases are detected when the JSON has `server_type` (typically `"vllm-omni"`) or `"diffusion"` in the `mark` array. Omni / TTS JSON has neither.
 
-**Running perf cases**
+#### Running perf cases
 
-Pass **`--test-config-file`** to run one JSON file as-is, or omit it for the bulk scan above and filter with pytest **`-m`** on each case's JSON `mark` (for example `pytest … run_benchmark.py -m "full_model and tts and H100"`).
+Pass **`--test-config-file`** to run one JSON file as-is, or omit it for the bulk scan above and filter with pytest **`-m`** on each case's JSON `mark` (for example `pytest … run_benchmark.py -m "full_model and tts and H100"` or `-m "slow and omni and H100"` for weekly Omni configs).
 
 ```JSON
 {
@@ -31,7 +32,7 @@ Pass **`--test-config-file`** to run one JSON file as-is, or omit it for the bul
     ],
     "server_params": {
         "model": "Qwen/Qwen3-Omni-30B-A3B-Instruct",
-        "stage_config_name": "qwen3_omni.yaml"
+        "stage_config_name": "qwen3_omni_moe.yaml"
     },
     "benchmark_params": [
         {
@@ -55,20 +56,22 @@ Pass **`--test-config-file`** to run one JSON file as-is, or omit it for the bul
 }
 ```
 
-**Parameter Explanation**
+#### Parameter Explanation
 
-*Overview*
+##### Overview
 
 | Field              | Required       | Description                                  |
-| ---------------- | -------- | --------------------------------------------------------------- |
+| ------------------ | -------------- | -------------------------------------------- |
 | test_name          | Yes            | Unique identifier for the test case          |
-| mark             | No       | Pytest marks for this case (see **`mark` field** below). Omit only for configs not meant to be filtered by `-m`. |
+| mark               | No             | Pytest marks; see **`mark` field** below     |
 | server_params      | Yes            | Server-side configuration parameters         |
-| benchmark_params | Yes      | Benchmark running parameters (supports multiple configurations) |
-| server_type      | Diffusion only | When set (e.g. `"vllm-omni"`), routes the case to `run_diffusion_benchmark.py`. |
-| benchmark_endpoint | Diffusion only | API path for the benchmark client (e.g. `/v1/videos`, `/v1/images/generations`). |
+| benchmark_params   | Yes            | Benchmark running parameters                 |
+| server_type        | Diffusion only | Routes case to run_diffusion_benchmark.py    |
+| benchmark_endpoint | Diffusion only | Benchmark API path                           |
 
-**`mark` field**
+Omit `mark` only for configs not meant to be filtered by `-m`. `server_type` is typically `"vllm-omni"`. `benchmark_endpoint` examples: `/v1/videos`, `/v1/images/generations`.
+
+#### `mark` field
 
 Optional top-level field on each perf JSON **case object** (one per `test_name`). `run_benchmark.py` and `run_diffusion_benchmark.py` read it via `tests.dfx.conftest.resolve_pytest_marks` and attach the marks to the corresponding `pytest.param`, so you can filter locally with `-m`.
 
@@ -119,11 +122,11 @@ Multi-GPU diffusion (example: Cosmos3 with `cfg-parallel-size=2`):
 - Diffusion perf: include `"diffusion"` in the `mark` array
 - HunyuanImage local-weight cases: add `"local_model"` to the `mark` array
 
-**Parametrization IDs**
+#### Parametrization IDs
 
 Each `(server, benchmark index)` pair becomes one `pytest.param` with id `{test_name}-{suffix}`. The suffix comes from `benchmark_params[].name` when set (for example `test_tts-p0`, `test_omni-p1`); otherwise it is derived from fields like `task` / `eval_phase`.
 
-**Benchmark result filenames**
+#### Benchmark result filenames
 
 Result files use the **runtime** hardware label from `get_runtime_resource_label()` (detected GPU/NPU name on the machine that ran the job), **not** `mark.hardware_marks.res`. On the default H100 CI pool, `H100` is **omitted** from filenames (`resource_label_for_filename`).
 
@@ -132,7 +135,7 @@ Examples:
 - Omni/TTS: `result_{test_name}_{optional_hw}_{dataset}_....json` under `BENCHMARK_DIR`
 - Diffusion: one aggregate `diffusion_result_{config_stem}_{optional_hw}_{timestamp}.json` per source JSON file (array of all runs from that file)
 
-**Local commands**
+#### Local commands
 
 ```bash
 # Bulk load + filter by JSON mark
@@ -146,18 +149,19 @@ pytest -s -v tests/dfx/perf/scripts/run_benchmark.py \
   --test-config-file tests/dfx/perf/tests/test_qwen3_omni_async_chunk.json
 ```
 
+<!-- markdownlint-disable-next-line MD051 -->
 See also [Markers for Tests](#markers-for-tests) for registered hardware markers (`H100`, `L4`, `cuda`, `distributed_cuda`, …).
 
-**`server_params` Configuration**
+#### `server_params` Configuration
 
-*Basic Parameters*
+##### Basic Parameters
 
 | Parameter         | Required | Example                   | Description                    |
-| ----------------- | -------- | ---------------------------------- | ----------------------------- |
-| model             | Yes      | "Qwen/Qwen3-Omni-30B-A3B-Instruct" | Model name or path            |
-| stage_config_name | Yes      | "qwen3_omni.yaml"                  | Stage configuration file name |
+| ----------------- | -------- | ------------------------- | ------------------------------ |
+| model             | Yes      | Qwen/Qwen3-Omni-30B-A3B…  | Model name or path             |
+| stage_config_name | Yes      | qwen3_omni_moe.yaml       | Deploy configuration file name |
 
-*Dynamic Configuration (update/delete)*
+##### Dynamic Configuration (update/delete)
 
 Supports incremental modifications based on the basic configuration:
 
@@ -166,25 +170,25 @@ Supports incremental modifications based on the basic configuration:
 | update    | Update or add configuration items    |
 | delete    | Delete specified configuration items |
 
-**Example**:
+##### Example
 
-```
+```json
 "update": {
     "async_chunk": true,  // Enable asynchronous chunk processing
-    "stage_args": {
+    "stages": {
         "0": {
-            "engine_args.custom_process_next_stage_input_func": "vllm_omni.model_executor.stage_input_processors.qwen3_omni.thinker2talker_async_chunk"
+            "max_num_seqs": 64
         }
     }
 },
 "delete": {
-    "stage_args": {
-        "2": ["custom_process_input_func"]  // Delete this configuration for stage 2
+    "stages": {
+        "2": ["max_num_batched_tokens"]  // Delete this deploy override for stage 2
     }
 }
 ```
 
-**`benchmark_params` Configuration**
+#### `benchmark_params` Configuration
 
 You can add any benchmark running parameters you need here. For all optional parameters, refer to the [benchmark documentation](https://github.com/vllm-project/vllm-omni/blob/main/docs/cli/bench/serve.md). General modifications are as follows:
 
@@ -195,9 +199,11 @@ You can add any benchmark running parameters you need here. For all optional par
 5. The qps and concurrency modes are recommended to be mutually exclusive. For detailed explanations, see the table below:
 
 | Parameter       | Type          | Required | Example/Values       | Description                          |
-| --------------- | ----------- | -------- | --------------- | -------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- |
-| name            | string      | No       | `"1024x1024_steps4"` | Stable suffix for pytest param id and logs. |
-| num_prompts     | int / array | Yes      | 10,[10, 20, 30] | Number of requests. Supports single values or arrays. If a single value is used, it will be automatically expanded to match the number of qps or max_concurrency, e.g., [10,10,10]. If an array is used, its length must match the number of qps or max_concurrency. |
-| request_rate    | float / array | No  | 0.5, [0.5, 1, inf] | Queries per second. Supports single values or arrays. If a single value is used, it will be automatically expanded to match the number of num_prompts, e.g., [1,1,1]. If an array is used, its length must match the number of num_prompts.                          |
-| max_concurrency | int / array | No       | 1, [1, 2, 3]    | Maximum concurrent in-flight requests. Same array / expansion rules as `request_rate` (mutually exclusive with QPS mode).                                                                                                                                                                                                             |
-| baseline        | object      | No       | see below       | Optional thresholds. **Hardware-nested only**: top-level keys must be `[hardware-resource]` pytest markers from `pyproject.toml` (`get_hardware_mark_list()`). Under each hardware key, metric names are free-form; values are scalars or sweep-aligned lists. Example: `{"H100": {"throughput_qps": […], "custom_stage_ms": 1.0}, "A3": {…}}`. Runtime filename aliases (`_RUNTIME_DEVICE_ALIASES`) are separate and do not expand this allowlist. |
+| --------------- | ------------- | -------- | -------------------- | ------------------------------------ |
+| name            | string        | No       | `"1024x1024_steps4"` | Stable suffix for pytest param id    |
+| num_prompts     | int / array   | Yes      | 10, [10, 20, 30]     | Number of requests (scalar or sweep) |
+| request_rate    | float / array | No       | 0.5, [0.5, 1, inf]   | Queries per second (QPS mode)        |
+| max_concurrency | int / array   | No       | 1, [1, 2, 3]         | Max in-flight requests               |
+| baseline        | object        | No       | see below            | Optional hardware-nested thresholds  |
+
+`num_prompts`, `request_rate`, and `max_concurrency` accept scalars or arrays; single values expand to match sweep length. QPS and concurrency modes are mutually exclusive. `baseline` top-level keys must be hardware pytest markers from `pyproject.toml`; metric names and values are free-form (scalars or sweep-aligned lists). Omit `baseline` to run without threshold assertions.

docs/contributing/ci/test_execution_guide.md

@@ -1,13 +1,20 @@
 # Test Execution Guide
+
 ## Setting Up the Test Environment
+
 ### Creating a Container
+
 vLLM-Omni provides an official Docker image for deployment. These images are built upon vLLM Docker images and are available on [Docker Hub](https://hub.docker.com/r/vllm/vllm-omni/tags). The version of vLLM-Omni indicates which vLLM release it is based on.
 For a local test environment, you can follow the steps below to create a container:
+
 ## Installing Dependencies
+
 ### vLLM & vLLM-Omni
+
 vLLM-Omni is built based on vLLM. You can follow [install guide](../../getting_started/installation/README.md) to build your local environment.
 
 ### Test Case Dependencies
+
 When running test cases, you may need to install the following dependencies:
 
 ```bash
@@ -16,6 +23,7 @@ apt-get install -y espeak-ng jq
 ```
 
 ## Running Tests
+
 Our test scripts use the pytest framework. First, please use `git clone https://github.com/vllm-project/vllm-omni.git` to download the vllm-omni source code. Then, in the root directory of vllm-omni, you can run the following commands in your local test environment to execute the corresponding test cases.
 
 ### CI job runners (L2–L4): logs, timing, and timeouts
@@ -25,7 +33,7 @@ Our test scripts use the pytest framework. First, please use `git clone https://
 **Log layout** (default under `logs/ready_jobs`, `logs/merge_jobs`, or `logs/nightly_jobs`; override with `--log-dir`):
 
 | Path | Description |
-|------|-------------|
+| ------ | ------------- |
 | `jobs/*.sh` | Generated per-step bash wrappers |
 | `jobs/.job_timeouts` | Job key → `timeout_in_minutes` from Buildkite YAML (when set) |
 | `<job_key>.log` | Tee output for each job |
@@ -33,7 +41,7 @@ Our test scripts use the pytest framework. First, please use `git clone https://
 
 **Timing:** while each job runs, stderr prints `finished in …` or `failed after …`. When all jobs finish, a summary is printed and written to `timing_summary.log`, for example:
 
-```
+```text
 === Job timing summary ===
   TTS_VoxCPM2_Test  25m 40s  OK
   Omni_Qwen3-Omni_Test  8m 05s  TIMED OUT
@@ -114,7 +122,6 @@ Failed jobs: 1/2
     pytest -s -v -m "advanced_model and distributed_cuda and L4" --run-level=advanced_model
     ```
 
-
 === "L4 level"
 
     **Recommended:** run CI-aligned nightly jobs from the repo root via [`tools/nightly/run_nightly_jobs.sh`](https://github.com/vllm-project/vllm-omni/blob/main/tools/nightly/run_nightly_jobs.sh). The script reads [`.buildkite/cuda/test-nightly.yml`](https://github.com/vllm-project/vllm-omni/blob/main/.buildkite/cuda/test-nightly.yml), generates per-step bash wrappers, runs pytest (perf jobs first, then aggregates perf JSON into Excel), and tees logs under `logs/nightly_jobs/` (see [CI job runners](#ci-job-runners-l2l4-logs-timing-and-timeouts); `generate_nightly_perf_excel` is included in the timing summary). Requires `bash`, `python3`, and PyYAML (`pip install pyyaml`).
@@ -160,7 +167,7 @@ Failed jobs: 1/2
     pytest -sv tests/dfx/perf/scripts/run_benchmark.py --test-config-file tests/dfx/perf/tests/test_tts.json
     pytest -sv tests/dfx/perf/scripts/run_diffusion_benchmark.py --test-config-file tests/dfx/perf/tests/test_cosmos3_vllm_omni.json
     ```
-    Nightly **Perf Test** jobs in ``test-nightly.yml`` use ``--test-config-file`` only (no ``-m``). E2e L4 function tests still use ``full_model`` + ``--run-level full_model`` (see [test-nightly.yml](https://github.com/vllm-project/vllm-omni/blob/main/.buildkite/test-nightly.yml)). Example:
+    Nightly **Perf Test** jobs in [``test-nightly.yml``](https://github.com/vllm-project/vllm-omni/blob/main/.buildkite/cuda/test-nightly.yml) use ``--test-config-file`` only (no ``-m``). Weekly **Perf Test** in [``test-weekly.yml``](https://github.com/vllm-project/vllm-omni/blob/main/.buildkite/cuda/test-weekly.yml) runs ``test_qwen3_omni_vllm_text.json`` and ``test_qwen3_omni_multi_replicas.json`` (JSON ``mark`` includes ``slow``). E2e L4 function tests use ``full_model`` + ``--run-level full_model``. Example:
 
 === "L5 level"
 
@@ -184,14 +191,15 @@ Failed jobs: 1/2
     # Reliability: HunyuanImage DiT (H100 × 4)
     pytest -s -v dfx/reliability/test_reliability_hunyuan_image.py -m slow
 
-    # Reliability: VoxCPM2 (L4)
-    pytest -s -v dfx/reliability/test_reliability_voxcpm2.py -m slow
-
     ```
 
-    The latest L5 CI jobs (reliability + invalid-parameter weekly steps) are in [test-weekly.yml](https://github.com/vllm-project/vllm-omni/blob/main/.buildkite/cuda/test-weekly.yml).
+    The latest L5 CI jobs are in [test-weekly.yml](https://github.com/vllm-project/vllm-omni/blob/main/.buildkite/cuda/test-weekly.yml):
+
+    - **Reliability** (+ invalid-parameter), **Perf Test** (Omni vLLM Text / Multi-Replica), and **Simple · CPU Coverage Test** → `WEEKLY=1` (or PR label `weekly-test` for Reliability/Perf). Scheduled upload of those env-gated steps still requires `main`.
+    - **E2E Tests** (slow Omni / TTS / Diffusion sweeps) → `NON_CRITICAL=1` (pipeline upload also accepts `WEEKLY=1` / `weekly-test`; the group `if` still needs `NON_CRITICAL`)
 
 You can find more information about markers in [Test Writing Guide — Markers for Tests](./test_writing_guide.md#markers-for-tests).
 
 ## Adding New Test Cases
+
 Please refer to the [Test System Overview](./test_system_overview.md) and [Test Writing Guide](./test_writing_guide.md).

docs/contributing/ci/test_system_overview.md

@@ -12,8 +12,7 @@ Through five levels (L1-L5) and common (Common) specifications, the system clari
       <th>Level</th>
       <th>Scope & Focus</th>
       <th>Model Coverage Strategy</th>
-      <th>Feature Coverage Strategy</th>
-      <th>Interface Coverage Strategy</th>
+      <th>Module Strategy</th>
       <th>Tags</th>
       <th>Time Cost</th>
       <th>Test Dir</th>
@@ -31,7 +30,6 @@ Through five levels (L1-L5) and common (Common) specifications, the system clari
       <td>/</td>
       <td>/</td>
       <td>/</td>
-      <td>/</td>
       <td><a href="https://github.com/vllm-project/vllm-omni/blob/main/.github/PULL_REQUEST_TEMPLATE.md">PR Checklist</a></td>
       <td>/</td>
       <td>/</td>
@@ -43,7 +41,6 @@ Through five levels (L1-L5) and common (Common) specifications, the system clari
       <td>/</td>
       <td>/</td>
       <td>/</td>
-      <td>/</td>
       <td><a href="./failures.md">CI Failures</a></td>
       <td>/</td>
       <td>/</td>
@@ -53,7 +50,6 @@ Through five levels (L1-L5) and common (Common) specifications, the system clari
       <td>Unit tests for components like entrypoints, models</td>
       <td>/</td>
       <td>/</td>
-      <td>/</td>
       <td><code>core_model and cpu</code></td>
       <td rowspan="2">&lt;15min</td>
       <td>
@@ -71,9 +67,8 @@ Through five levels (L1-L5) and common (Common) specifications, the system clari
     <tr>
       <td><strong>L2</strong><br>(E2E across models & GPU-required UT)</td>
       <td>Online (basic deployment scenarios):<br>dummy, normal inference function (output format, stream), some instance startup UT</td>
-      <td>High-priority models + online basic scenarios; request success, non-empty output, format match (no Whisper/accuracy)</td>
-      <td>High-priority features and component tests that require launching instances</td>
-      <td>High-priority interfaces (using random lightweight models)</td>
+      <td>Key models + online basic scenarios; request success, non-empty output, format match (no Whisper/accuracy)</td>
+      <td>Key features and modules that require launching instances (prefer random weights)</td>
       <td><code>core_model and hardware_test(H100, L4, etc.) and omni/tts/diffusion</code></td>
       <td>
         <strong>Model E2E:</strong><br>
@@ -95,9 +90,8 @@ Through five levels (L1-L5) and common (Common) specifications, the system clari
     <tr>
       <td><strong>L3</strong><br>(Important Perf & Integration & Accuracy)</td>
       <td>Online & Offline (multiple deployment scenarios):<br>real model, normal inference function, normal accuracy</td>
-      <td>High/medium-priority models + key online/offline scenarios; real weights, Whisper/similarity, preset voice gender, basic accuracy</td>
-      <td>Medium-priority features and component tests that require launching instances</td>
-      <td>Medium-priority interfaces (using random lightweight models)</td>
+      <td>Key models + key online/offline scenarios; real weights, Whisper/similarity, preset voice gender, basic accuracy</td>
+      <td>Key features and modules that require launching instances (using real weights)</td>
       <td><code>advanced_model and hardware_test(H100, L4, etc.) and omni/tts/diffusion</code></td>
       <td>&lt;30min</td>
       <td>
@@ -120,9 +114,8 @@ Through five levels (L1-L5) and common (Common) specifications, the system clari
     <tr>
       <td><strong>L4</strong><br>(Perf & Integration & Accuracy)</td>
       <td>Online: full functional scenarios + performance test + doc test + accuracy test</td>
-      <td>High-priority models: function, performance, accuracy, and doc testing<br>Medium-priority models: function and doc testing</td>
-      <td>Low-priority features and component tests that require launching instances</td>
-      <td>Low-priority interfaces (using real weights)</td>
+      <td>Key models: function, performance, accuracy, and doc testing</td>
+      <td>Other features and modules that require launching instances (using real weights)</td>
       <td><code>full_model and hardware_test(H100, L4, etc.) and omni/tts/diffusion</code></td>
       <td>&lt;3 hour</td>
       <td>
@@ -147,11 +140,10 @@ Through five levels (L1-L5) and common (Common) specifications, the system clari
       <td>GPU</td>
     </tr>
     <tr>
-      <td><strong>L5</strong><br>(Stability & Reliability)</td>
-      <td>Online: long-term stability test + reliability test</td>
-      <td>Long-term stability and reliability testing for high-priority models<br>Low-priority models: function and doc testing</td>
+      <td><strong>L5</strong><br>(Stability & Reliability & selected Perf)</td>
+      <td>Online: long-term stability + reliability; coverage test</td>
+      <td>Long-term stability and reliability testing for key models<br>Non-critical scenario performance and accuracy testing for key models<br>Other models: function and doc testing</td>
       <td>/</td>
-      <td>Invalid-parameter validation for high-priority interfaces</td>
       <td><code>slow and hardware_test(H100, L4, etc.) and omni/tts/diffusion</code></td>
       <td> Depends on reality </td>
       <td>

docs/contributing/ci/test_writing_guide.md

@@ -17,17 +17,18 @@ Summary:
 - **Reliability:** `tests/dfx/reliability/`
 - Do **not** add new top-level `tests/` directories unrelated to a component (or to `e2e` / `dfx` / `helpers` / `examples` / `buildkite`)
 
-Deploy YAMLs: `vllm_omni/deploy/` via `tests.helpers.stage_config.get_deploy_config_path` (see [stage configs](../../configuration/stage_configs.md)).
+Deploy YAMLs: `vllm_omni/deploy/` via `tests.helpers.stage_config.get_deploy_config_path` (see [pipeline and deploy configs](../../configuration/stage_configs.md)).
 
 ## Markers for Tests
 
 By adding markers before test functions, tests can later be executed uniformly by simply declaring the corresponding marker type.
 
 ### Current Markers
+
 Defined in `pyproject.toml`:
 
 | Marker             | Description                                                                                             |
-| ------------------ | --------------------------------------------------------- |
+| ------------------ | ------------------------------------------------------------------------------------------------------- |
 | `core_model`       | L1&L2 tests (run in each PR)                                                                            |
 | `advanced_model`   | L3 tests (run on each merge to main)                                                                    |
 | `full_model`       | L4 tests (run nightly)                                                                                  |
@@ -81,13 +82,13 @@ This decorator is intended to make hardware-aware, cross-platform test authoring
 
 1. **Applies platform and resource markers**  
    Adds the appropriate pytest markers for each specified hardware platform (e.g., `cuda`, `rocm`, `xpu`, `npu`) and resource type (e.g., `L4`, `H100`, `MI325`, `B60`, `A2`, `A3`).
-   ```
+   ```python
    @pytest.mark.cuda
    @pytest.mark.L4
    ```
 2. **Handles multi-card (distributed) scenarios**  
    For tests requiring multiple cards, it automatically adds distributed markers such as `distributed_cuda`, `distributed_rocm`, or `distributed_npu`.
-   ```
+   ```python
    @pytest.mark.distributed_cuda(num_cards=num_cards)
    ```
 3. **Supports flexible card requirements**  
@@ -97,11 +98,11 @@ This decorator is intended to make hardware-aware, cross-platform test authoring
    On CUDA, adds a skip marker (`skipif_cuda`) if the system does not have the required number of devices.
    Support for `skipif_rocm` and `skipif_npu` will be implemented later.
 
-
 5. **Works with pytest filtering**  
    Allows tests to be filtered and selected at runtime using standard pytest marker expressions (e.g., `-m "distributed_cuda and L4"`).
 
 ##### Example usage for decorator
+
 - Single call for multiple platforms:
     ```python
     @hardware_test(
@@ -220,6 +221,7 @@ If you want to add support for a new platform (e.g., "tpu" for a new accelerator
    ```
 
 **Summary**:  
+
 - Add pytest markers for your new platform/resources  
 - Implement a marker function (`xxx_marks`)  
 - Plug into `hardware_marks`  
@@ -237,16 +239,14 @@ L1 and L2 level testing form the foundation of the quality assurance system. L1
 
 L2 level testing builds upon L1 by introducing GPU resources and verifying that the end-to-end (E2E) process of the model in basic deployment scenarios is smooth. For example, it uses dummy models to confirm that core interfaces like the inference pipeline, output format, and streaming response work properly. The common goal of these two levels is to provide developers with rapid feedback, discovering and fixing issues early in the development cycle.
 
-
-
 #### 1.2 Testing Content and Scope
 
 - ***L1 (Unit & Logic Testing)***:
--   -   ***Scope***: Tests internal functions and methods of core components such as `entrypoints`, `models`.
+    - ***Scope***: Tests internal functions and methods of core components such as `entrypoints`, `models`.
         - ***Focus***: Branch coverage, exception handling, algorithm logic correctness. Does not involve external dependencies or the complete service stack.
         - ***Time Cost***: Execution time is controlled within ***15 minutes*** to ensure fast feedback.
 - ***L2 (Basic End-to-End Testing)***:
--   -   ***Scope***: Covers two basic deployment scenarios: `online` (serving) and `offline` (inference).
+    - ***Scope***: Covers two basic deployment scenarios: `online` (serving) and `offline` (inference).
         - ***Focus***: Uses `dummy` weights (via deploy YAML patching at `core_model`) or lightweight real models to verify that the entire chain from request input to result output works normally, including output data structure, streaming (stream) support, and **cheap payload checks** at `--run-level core_model` (see below). Also includes some unit tests that require launching independent service instances.
         - ***L2 response validation (`core_model`)***: Implemented in `tests/helpers/assertions.py` and invoked by `OpenAIClientHandler` based on `--run-level`. At L2 we require **request success** plus minimal output sanity—not full accuracy:
             - **Speech / TTS** (`assert_audio_speech_response`): decoded audio must be present and non-empty (or exceed `min_audio_bytes` when set in `request_config`); `response_format` must match the returned content-type (e.g. `wav`, `pcm`). Whisper transcript similarity, PCM HNR, and preset-voice gender checks run only at L3+.
@@ -275,9 +275,10 @@ A clear directory structure is key to managing test cases efficiently. See the *
 <summary> L1 Test Examples</summary>
 
 Examples from `tests/model_executor/models/qwen2_5_omni/test_audio_length.py`
+
 ```python
 # SPDX-License-Identifier: Apache-2.0
-# SPDX-FileCopyrightText: Copyright contributors to the vLLM project
+# SPDX-FileCopyrightText: Copyright contributors to the vLLM-Omni project
 
 import pytest
 
@@ -331,6 +332,7 @@ def test_cap_and_align_mel_length_no_mismatch(repeats, code_len, max_mel_frames)
     if max_mel_frames is not None and int(max_mel_frames) > 0 and int(max_mel_frames) >= repeats:
         assert target_mel_len <= int(max_mel_frames)
 ```
+
 </details>
 
 <details>
@@ -338,7 +340,7 @@ def test_cap_and_align_mel_length_no_mismatch(repeats, code_len, max_mel_frames)
 You can refer to Test Examples in the L3 section to see example test cases that incorporate both L2 and L3 testing logic.
 </details>
 
--   -   ***Run Command***:
+- ***Run Command***:
 
     `pytest -s -v /tests/e2e/online_serving/test_{model_name}.py`
     `pytest -s -v -m 'core_model and cpu' --run-level=core_model`
@@ -351,13 +353,11 @@ L3 level testing executes after code is merged into the main branch. Its core pu
 
 . It acts as the "quality gatekeeper" for the main branch, ensuring that no merge breaks the core capabilities of the model service. Testing needs to provide clear conclusions within a relatively short time (<30min), balancing test depth with feedback speed.
 
-
-
 #### 2.2 Testing Content and Scope
 
 - ***Deployment Scenarios***: Covers richer `online` and `offline` deployment configurations, which may include different hardware configurations, batch sizes, concurrency levels, etc.
 - ***Core Verification***:
--   1.  ***Inference Functionality***: Ensures real models can perform forward computation normally and return results.
+    1. ***Inference Functionality***: Ensures real models can perform forward computation normally and return results.
     2. ***Accuracy Compliance***: Verifies that the model's evaluation metrics (e.g., accuracy) meet the expected baseline, preventing code changes from introducing accuracy issues.
     3. ***Important Performance***: Verifies whether performance (e.g., P99 latency, throughput) in core scenarios meets preset thresholds.
 - ***L3 response validation (`advanced_model`)***: At `--run-level advanced_model`, `tests/helpers/assertions.py` adds semantic checks on top of L2 payload gates (see L1 & L2 above):
@@ -506,12 +506,10 @@ L3 level testing executes after code is merged into the main branch. Its core pu
 
 L4 level testing is a comprehensive quality audit before a version release. It expands upon L3, executing ***full*** functional scenarios, conducting systematic ***performance stress tests***, and simultaneously verifying the correctness of accompanying ***example documentation***. Its purpose is to perform deep validation of the system during off-peak nighttime hours, providing quality trend reports for daytime development and data support for release decisions.
 
-
-
 #### 3.2 Testing Content and Scope
 
 - ***Full Functionality Testing***: Executes all test cases defined in `test_{model_name}_expansion.py`, covering all implemented features, positive flows, boundary conditions, and exception handling.
--   ***Performance Testing***: Uses `tests/dfx/perf/tests/test_qwen3_omni_*.json` (Omni), `test_tts.json` / `test_voxcpm2.json` / `test_higgs_audio_v3.json` (TTS), and diffusion configs `tests/dfx/perf/tests/test_*_vllm_omni.json` (passed to `run_benchmark.py` or `run_diffusion_benchmark.py` via `--test-config-file` in nightly **Perf Test** steps) to drive throughput, latency, and memory benchmarks. Each JSON **case** may declare an optional top-level **`mark`** array: exactly one ``hardware_marks`` object plus pytest marker name strings (`full_model`, `omni` / `tts` / `diffusion`, …). Runners attach those marks to each parametrized `(server, benchmark index)` pair so **local** bulk runs can filter with `-m` (for example `-m "full_model and H100 and diffusion"`). Nightly CI perf jobs select workloads by **`--test-config-file`**, not `-m`. Details are in the Performance Tests example below.
+- ***Performance Testing***: JSON configs under `tests/dfx/perf/tests/` drive Omni, TTS, and diffusion serving benchmarks via `run_benchmark.py` or `run_diffusion_benchmark.py`, measuring throughput, latency, and related metrics under configured workloads.
 - ***Documentation Testing***: Verifies whether the example code provided to users is runnable and its results match the description.
 
 #### 3.3 Test Directory and Execution Files
@@ -548,12 +546,10 @@ L4 level testing is a comprehensive quality audit before a version release. It e
 
 L5 level testing focuses on the performance of model services under ***long-running*** and ***abnormal fault*** scenarios. It aims to uncover deep-seated issues that only manifest under sustained pressure or extreme conditions, such as memory leaks, resource contention, gradual performance degradation, and lack of fault tolerance mechanisms. This is the final, yet crucial, line of defense for ensuring service high availability and production environment robustness.
 
-
-
 #### 4.2 Testing Content and Scope
 
 - ***Long-term Stability (Stability) Testing***: Uses JSON under `tests/dfx/stability/tests/` (for example `test_qwen3_omni.json` and `test_wan22.json`) to run the service under moderate load for an extended period (e.g., over 12 hours), monitoring whether metrics like memory/VRAM usage, response time, and throughput degrade over time, and whether the service process remains stable.
--   ***Reliability Testing***: Uses pytest suites under `tests/dfx/reliability/` to inject controlled faults against a **live** `vllm_omni serve` instance (same **`omni_server` / `omni_server_function`** fixture style as E2E). Current suites emphasize **GPU memory pressure** (CUDA sidecar “memory hog”), **worker / runtime process kill** (`SIGKILL` on `VLLM::Worker` for Qwen3-Omni, `multiprocessing.spawn` for Wan2.2 video workers, or `vLLM-Omni::` for HunyuanImage DiT workers), **large multimodal chat**, **`/v1/videos`**, **`/v1/images/generations`**, or **`/v1/audio/speech`** jobs under OOM, **`/health` → 503** and **fast-fail / non-hanging concurrent** requests after kill, and **OpenAI-style 5xx error contracts** (e.g. text vs text+audio under OOM). **Post-fault recovery** checks exist where enabled (some cases may be `skip` while issues are tracked). See the Reliability `<details>` block in Section 4.4 for file-level responsibilities and CI markers (`slow`, `hardware_test`, POSIX-only kill).
+- ***Reliability Testing***: Uses pytest suites under `tests/dfx/reliability/` to inject controlled faults against a **live** `vllm_omni serve` instance (same **`omni_server` / `omni_server_function`** fixture style as E2E). Current suites emphasize **GPU memory pressure** (CUDA sidecar “memory hog”), **worker / runtime process kill** (`SIGKILL` on `VLLM::Worker` for Qwen3-Omni, `multiprocessing.spawn` for Wan2.2 video workers, or `vLLM-Omni::` for HunyuanImage DiT workers), **large multimodal chat**, **`/v1/videos`**, **`/v1/images/generations`**, or **`/v1/audio/speech`** jobs under OOM, **`/health` → 503** and **fast-fail / non-hanging concurrent** requests after kill, and **OpenAI-style 5xx error contracts** (e.g. text vs text+audio under OOM). **Post-fault recovery** checks exist where enabled (some cases may be `skip` while issues are tracked). See Section 4.3 for file-level responsibilities and Section 4.4 Test Examples (Reliability) for CI markers (`slow`, `hardware_test`, POSIX-only kill).
 
 #### 4.3 Test Directory and Execution Files
 
@@ -562,7 +558,6 @@ L5 level testing focuses on the performance of model services under ***long-runn
     - `test_reliability_qwen3_omni.py` — Qwen3-Omni chat / multimodal reliability (GPU OOM, process kill, recovery, error contract under `--async-chunk` vs default).
     - `test_reliability_wan22.py` — Wan2.2 I2V video API reliability (`/v1/videos` under OOM and process kill, recovery).
     - `test_reliability_hunyuan_image.py` — HunyuanImage-3.0-Instruct DiT-only reliability (`/v1/images/generations` under OOM and process kill; deploy `hunyuan_image3_dit.yaml`, H100 × 4).
-    -   `test_reliability_voxcpm2.py` — VoxCPM2 TTS reliability (`/v1/audio/speech` under OOM and process kill; deploy `voxcpm2.yaml`, L4 × 1).
     - `helpers.py` — Shared primitives used by current suites: raw HTTP probes for `/v1/chat/completions` and `/health`, OpenAI-style error parsing, GPU OOM sidecar (`inject_gpu_oom` / `stop_gpu_oom_hogs`), and `pgrep`-based process-kill injector construction (`make_process_kill_fault_injector`).
     - `conftest.py` — `fault_injector` and `omni_server_after_fault` / `omni_server_after_fault_function` fixtures to run a callable **after** the server is ready.
     - `README.md` — Short local run commands for this directory.
@@ -570,19 +565,24 @@ L5 level testing focuses on the performance of model services under ***long-runn
 #### 4.4 Execution Method and Example
 
 - ***Trigger Timing***: **`Weekly`** (weekly) or **`Days before Release`** (several days before a major release). Due to long execution times, the frequency is lower.
--   ***Execution Environment***: ***GPU*** servers, requiring a stable and exclusive testing environment.
+- ***Run Command***:
+    - ***Stability***: `pytest -s -v tests/dfx/stability/scripts/test_stability_qwen3_omni.py` or `pytest -s -v tests/dfx/stability/scripts/test_stability_wan22.py` (or add `test_stability_<model>.py` alongside a matching JSON config)
+    - ***Reliability***: `pytest -s -v tests/dfx/reliability/test_reliability_<model>.py -m slow` (current suites: `qwen3_omni`, `wan22`, `hunyuan_image`).
 - ***Script Example***:
+
 <details>
 <summary> Test Examples</summary>
 
-When you want to add L5-level stability test cases, add or extend the appropriate JSON file under `tests/dfx/stability/tests/` (for example `test_qwen3_omni.json` for Omni bench traffic, or `test_wan22.json` for diffusion `/v1/videos` workloads). The following illustrates the Qwen3-Omni shape:
+##### Stability
+
+When you want to add L5-level stability test cases, add or extend the appropriate JSON file under `tests/dfx/stability/tests/` (for example `test_qwen3_omni.json` for Omni bench traffic, or `test_wan22.json` for diffusion `/v1/videos` workloads). Pair the JSON with `tests/dfx/stability/scripts/test_stability_<model>.py`. The following illustrates the Qwen3-Omni shape:
 
 ```json
 {
     "test_name": "test_qwen3_omni_stability",
     "server_params": {
         "model": "Qwen/Qwen3-Omni-30B-A3B-Instruct",
-        "stage_config_name": "qwen3_omni.yaml"
+        "stage_config_name": "qwen3_omni_moe.yaml"
     },
     "benchmark_params": [
         {
@@ -601,38 +601,35 @@ When you want to add L5-level stability test cases, add or extend the appropriat
 }
 ```
 
-##### Parameter Explanation
+###### Parameter Explanation
 
 ***Overview***
 
 | Field            | Required | Description                                                                 |
 | ---------------- | -------- | --------------------------------------------------------------------------- |
 | test_name        | Yes      | Unique identifier for the stability test case                               |
-| server_params    | Yes      | Server-side configuration parameters (model, stage configuration, etc.)     |
+| server_params    | Yes      | Server-side configuration parameters (model, deploy configuration, etc.)    |
 | benchmark_params | Yes      | Stability benchmark running parameters (supports multiple configurations)   |
 
-##### server_params Configuration
+###### server_params Configuration
 
-###### Basic Parameters
+Basic parameters:
 
 | Parameter         | Required | Example                            | Description                         |
 | ----------------- | -------- | ---------------------------------- | ----------------------------------- |
 | model             | Yes      | "Qwen/Qwen3-Omni-30B-A3B-Instruct" | Model name or path                  |
-| stage_config_name | Yes      | "qwen3_omni.yaml"                  | Stage configuration file name       |
+| stage_config_name | Yes      | "qwen3_omni_moe.yaml"              | Deploy configuration file name      |
 
-###### Dynamic Configuration (update/delete)
-
-Supports incremental modifications based on the basic configuration:
+Dynamic configuration (`update` / `delete`) supports incremental modifications based on the basic configuration:
 
 | Operation | Description                          |
 | --------- | ------------------------------------ |
 | update    | Update or add configuration items    |
 | delete    | Delete specified configuration items |
 
-***Example***:
-You can refer to Test Examples in L4 §3.4
+***Example***: You can refer to Test Examples in L4 §3.4.
 
-##### benchmark_params Configuration
+###### benchmark_params Configuration
 
 For stability testing, the key parameters are:
 
@@ -644,69 +641,21 @@ For stability testing, the key parameters are:
 
 All other optional parameters follow the same rules as in L4 §3.4.
 
-</details>
+##### Reliability
 
-<details>
-<summary> Reliability test suite (<code>tests/dfx/reliability</code>)</summary>
+Add pytest modules under `tests/dfx/reliability/` (for example `test_reliability_qwen3_omni.py`, `test_reliability_wan22.py`, `test_reliability_hunyuan_image.py`). Reuse `helpers.py` and `conftest.py` rather than duplicating OOM / process-kill / raw HTTP plumbing.
 
-##### Purpose and relationship to stability
-
-Reliability tests are **short fault-injection** integration runs (L5 **(b)** in `tests/dfx/reliability/README.md`). They complement **stability** JSON-driven long runs: instead of hours of steady traffic, they **perturb** the server (GPU OOM sidecar, fatal signals on selected processes) and check **failure mode** and **latency bounds** (e.g. chat or `/v1/videos` must not hang under concurrent fault-time load).
-
-##### Directory layout
-
-| Path | Responsibility |
-| ---- | -------------- |
-| `helpers.py` | Shared helpers used by current reliability suites: raw `POST`/`GET` probes (`/v1/chat/completions`, `/health`), OpenAI error parsing (`extract_openai_error_contract_from_bytes`), GPU OOM sidecar lifecycle (`inject_gpu_oom`, `stop_gpu_oom_hogs`), and process-kill injector builder (`make_process_kill_fault_injector`). |
-| `conftest.py` | Pytest fixtures: indirect `fault_injector`, `omni_server_after_fault` / `omni_server_after_fault_function` (run injector after server is ready, then yield server). |
-| `test_reliability_qwen3_omni.py` | Qwen3-Omni: OOM vs **text vs text+audio** error contract, large multimodal chat under OOM, concurrent pressure, **SIGKILL** on `VLLM::Worker`, `/health` → 503 + fast-fail + concurrent chat; optional OOM recovery scenario (may be skipped while tracked in issues). |
-| `test_reliability_wan22.py` | Wan2.2 I2V: large `/v1/videos` under OOM, **SIGKILL** on `multiprocessing.spawn` chain, health / fast-fail / concurrent video requests; optional recovery test (may be skipped). |
-| `test_reliability_hunyuan_image.py` | HunyuanImage DiT-only: large `/v1/images/generations` under OOM, **SIGKILL** on `vLLM-Omni::` workers and serve/tree targets, health / fast-fail / concurrent image requests; some OOM/recovery cases may be skipped while tracked in issues. |
-| `test_reliability_voxcpm2.py` | VoxCPM2: `/v1/audio/speech` under OOM (error contract), **SIGKILL** on `VLLM::` workers and serve/tree targets, health / fast-fail / concurrent speech requests; some OOM cases may be skipped while tracked in issues. |
-| `README.md` | Minimal run / collect examples. |
-
-##### Parametrization and markers
+###### Parametrization and markers
 
 - Each test module defines a **`RELIABILITY_SCENARIOS`** list (`test_name`, `server_params`: model, `stage_config_name` or diffusion `server_args`, etc.). **`create_reliability_omni_server_params()`** in `tests/dfx/conftest.py` resolves stage paths (including XPU substitutions where applicable) and builds **`OmniServerParams`** lists consumed by **`@pytest.mark.parametrize(..., indirect=True)`** on `omni_server` or `omni_server_function`.
-- Cases are tagged **`@pytest.mark.slow`** for weekly / selective CI. GPU-heavy suites use **`@hardware_test(res={"cuda": "H100"}, num_cards=...)`** or **`@hardware_test(res={"cuda": "L4"}, num_cards=1)`** (Qwen3-Omni **2**× H100; Wan2.2 **1**× H100; HunyuanImage DiT **4**× H100; VoxCPM2 **1**× L4).
+- Cases are tagged **`@pytest.mark.slow`** for weekly / selective CI. GPU-heavy suites use **`@hardware_test(res={"cuda": "H100"}, num_cards=...)`** (Qwen3-Omni **2**× H100; Wan2.2 **1**× H100; HunyuanImage DiT **4**× H100).
 - **Process-kill** tests use **`@pytest.mark.skipif(os.name == "nt", ...)`** because injection uses POSIX **`pgrep` / `kill`**.
 
-##### CI trigger
-
-Weekly Buildkite (`.buildkite/cuda/test-weekly.yml`) runs one step per model suite (trigger: `WEEKLY=1` or PR label `weekly-test`), for example:
-
-| Buildkite step | Test file | CI hardware |
-| -------------- | --------- | ----------- |
-| Reliability Test - qwen3-omni | `test_reliability_qwen3_omni.py` | H100 × 2 (`mithril-h100-pool`) |
-| Reliability Test - wan22 | `test_reliability_wan22.py` | H100 × 2 (`mithril-h100-pool`) |
-| Reliability Test - hunyuan-image | `test_reliability_hunyuan_image.py` | H100 × 4 (`mithril-h100-pool`) |
-| Reliability Test - voxcpm2 | `test_reliability_voxcpm2.py` | L4 × 1 (`gpu_1_queue`) |
-
-```bash
-pytest -s -v tests/dfx/reliability/test_reliability_qwen3_omni.py -m "slow"
-pytest -s -v tests/dfx/reliability/test_reliability_wan22.py -m "slow"
-pytest -s -v tests/dfx/reliability/test_reliability_hunyuan_image.py -m "slow"
-pytest -s -v tests/dfx/reliability/test_reliability_voxcpm2.py -m "slow"
-```
-
-##### Local commands
-
-```bash
-pytest --collect-only tests/dfx/reliability
-pytest -s -v tests/dfx/reliability/test_reliability_qwen3_omni.py -m slow
-pytest -s -v tests/dfx/reliability/test_reliability_wan22.py -m slow
-pytest -s -v tests/dfx/reliability/test_reliability_hunyuan_image.py -m slow
-pytest -s -v tests/dfx/reliability/test_reliability_voxcpm2.py -m slow
-```
-
-##### Adding a new model suite
+###### Adding a new model suite
 
 1. Add `test_reliability_<model>.py` under `tests/dfx/reliability/`.
-2. Define **`RELIABILITY_SCENARIOS`** and pass them through **`create_reliability_omni_server_params()`** with the correct deploy or e2e stage-config directory (same pattern as existing files).
-3. Reuse **`helpers`** for OOM / kill / raw HTTP; prefer **`assert_fault_exception()`** and **`resolve_oom_device_spec()`** from `tests/dfx/conftest.py` for consistent device selection vs stage YAML.
+2. Define **`RELIABILITY_SCENARIOS`** and pass them through **`create_reliability_omni_server_params()`** with the correct deploy or e2e deploy-config directory (same pattern as existing files).
+3. Reuse **`helpers`** for OOM / kill / raw HTTP; prefer **`assert_fault_exception()`** and **`resolve_oom_device_spec()`** from `tests/dfx/conftest.py` for consistent device selection vs deploy YAML.
 4. Register **`slow`** (and **`hardware_test`** if needed); extend **`.buildkite/cuda/test-weekly.yml`** when the suite should run in weekly L5.
 
 </details>
-
--   -   ***Stability***: `pytest -s -v tests/dfx/stability/scripts/test_stability_qwen3_omni.py` or `pytest -s -v tests/dfx/stability/scripts/test_stability_wan22.py` (or add `test_stability_<model>.py` alongside a matching JSON config)
-    -   ***Reliability***: `pytest -s -v tests/dfx/reliability/test_reliability_<model>.py -m slow` (current suites: `qwen3_omni`, `wan22`, `hunyuan_image`, `voxcpm2`; add `test_reliability_<suite>.py` and a matching step in `.buildkite/cuda/test-weekly.yml` for new models)

docs/contributing/model/adding_omni_model.md

@@ -9,7 +9,7 @@ This guide walks through the process of adding a new multi-stage model to vLLM-O
 3. [Step-by-Step Implementation](#step-by-step-implementation)
 4. [Key Components](#key-components)
 5. [Model Registration](#model-registration)
-6. [Stage Configuration](#stage-configuration)
+6. [Pipeline and Deploy Configuration](#pipeline-and-deploy-configuration)
 7. [Stage Input Processors](#stage-input-processors)
 8. [Testing](#testing)
 9. [Adding a Model Recipe](#adding-a-model-recipe)
@@ -316,18 +316,28 @@ _OMNI_MODELS = {
 
 The registry uses lazy loading, so the model class is imported only when needed.
 
-## Stage Configuration
+## Pipeline and Deploy Configuration
 
-Create a YAML configuration file in `vllm_omni/deploy/`. For a complete example, see the [Qwen3-Omni configuration file](gh-file:vllm_omni/deploy/qwen3_omni_moe.yaml).
+Define fixed topology in `vllm_omni/model_executor/models/<model>/pipeline.py`
+with `PipelineConfig` and `StagePipelineConfig`, then register it in
+`vllm_omni/config/pipeline_registry.py`. For a complete example, see the
+[Qwen3-Omni pipeline](gh-file:vllm_omni/model_executor/models/qwen3_omni/pipeline.py).
 
-### Key Configuration Fields
+### Pipeline topology fields
 
-- **`model_stage`**: Which stage to run ("thinker", "talker", "code2wav", etc.)
-- **`model_arch`**: The model architecture name (must match registry)
-- **`engine_input_source`**: List of stage IDs that provide input to this stage
-- **`custom_process_input_func`**: Function to process inputs from previous stages
-- **`final_output`**: Whether this stage produces the final output (True/False)
-- **`final_output_type`**: Type of final output ("text", "audio", "image", etc.)
+- **`model_stage`**: Logical stage name ("thinker", "talker", "code2wav", etc.)
+- **`execution_type`**: Runtime family (`LLM_AR`, `LLM_GENERATION`, or `DIFFUSION`)
+- **`model_arch`**: Stage architecture name, when different from the pipeline default
+- **`input_sources`**: Upstream stage IDs that provide input to this stage
+- **`custom_process_input_func`**: Function that processes inputs from previous stages
+- **`final_output`**: Whether this stage produces user-facing output
+- **`final_output_type`**: Output modality ("text", "audio", "image", etc.)
+- **`owns_tokenizer`**: Whether this stage owns the pipeline tokenizer
+
+Create the matching runtime defaults in `vllm_omni/deploy/<model>.yaml` and set
+`PipelineConfig.default_deploy_config_name` to that file. The deploy YAML owns
+GPU placement, memory sizing, parallelism, connectors, and sampling defaults;
+it must not redefine topology fields.
 
 ## Stage Input Processors
 
@@ -374,7 +384,7 @@ The stage transition process follows these steps:
 
 2. **Transition Detection**: The orchestrator checks if there's a next stage and calls `process_engine_inputs()` on it
 
-3. **Input Processing**: The stage input processor configured in stage YAML (under `vllm_omni/model_executor/stage_input_processors/`) handles the transition:
+3. **Input Processing**: The stage input processor registered by the pipeline (implemented under `vllm_omni/model_executor/stage_input_processors/`) handles the transition:
    ```python
    def process_engine_inputs(
        self, stage_list: list[Any], prompt: OmniTokensPrompt | TextPrompt = None
@@ -577,7 +587,7 @@ Adding a new model to vLLM-Omni involves:
 1. **Create model directory structure** with stage implementations
 2. **Implement unified model class** that orchestrates stages
 3. **Register model** in `registry.py`
-4. **Create stage configuration** YAML file
+4. **Define pipeline topology and deploy defaults** in `pipeline.py` and `vllm_omni/deploy/`
 5. **Implement stage input processors** for stage transitions
 6. **Write tests** to verify functionality
 7. **Add model recipe** to the [vllm-project/recipes](https://github.com/vllm-project/recipes) repository (see [Adding a Model Recipe](#adding-a-model-recipe) section)
@@ -590,7 +600,8 @@ For a complete reference implementation, see:
 - **Thinker**: `vllm_omni/model_executor/models/qwen3_omni/qwen3_omni_moe_thinker.py`
 - **Talker**: `vllm_omni/model_executor/models/qwen3_omni/qwen3_omni_moe_talker.py`
 - **Code2Wav**: `vllm_omni/model_executor/models/qwen3_omni/qwen3_omni_code2wav.py`
-- **Stage config**: `vllm_omni/deploy/qwen3_omni_moe.yaml`
+- **Pipeline topology**: `vllm_omni/model_executor/models/qwen3_omni/pipeline.py`
+- **Deploy config**: `vllm_omni/deploy/qwen3_omni_moe.yaml`
 - **Input processors**: `vllm_omni/model_executor/stage_input_processors/qwen3_omni.py`
 - **Registry**: `vllm_omni/model_executor/models/registry.py`
 - **Testing**: `vllm_omni/tests/e2e/offline_inference/test_qwen3_omni.py`
@@ -598,4 +609,4 @@ For a complete reference implementation, see:
 For more information, see:
 - [Architecture Overview](../../design/architecture_overview.md)
 - [Supported Models](../../models/supported_models.md)
-- [Stage Configuration Guide](../../configuration/stage_configs.md)
+- [Pipeline and Deploy Configuration Guide](../../configuration/stage_configs.md)

docs/contributing/model/adding_tts_model.md

@@ -20,7 +20,7 @@ pattern, refer to MOSS-TTS-Nano.
 4. [Step-by-Step Implementation](#step-by-step-implementation)
 5. [Key Components](#key-components)
 6. [Model Registration](#model-registration)
-7. [Stage Configuration](#stage-configuration)
+7. [Pipeline and Deploy Configuration](#pipeline-and-deploy-configuration)
 8. [Stage Input Processors](#stage-input-processors)
 9. [Online Serving Integration](#online-serving-integration)
 10. [Single-Stage Models](#single-stage-models)
@@ -78,7 +78,7 @@ Profile before optimizing.
 new TTS model must pass all three levels:
 
 | Layer | Catches | Tool |
-|-------|---------|------|
+| ------- | --------- | ------ |
 | Offline RTF / duration | Throughput regressions, missing audio, wrong sample rate | `end2end.py`, pytest e2e |
 | Browser streaming playback | Delta-vs-cumulative bugs, chunk boundary glitches, TTFP regressions | Gradio demo over `/v1/audio/speech?stream=true` |
 | Concurrent requests | Per-request state leaks, codec window round-robin gaps | `max_num_seqs>1` smoke with 4+ parallel prompts |
@@ -95,7 +95,7 @@ vLLM-Omni supports TTS models as multi-stage pipelines where each stage runs ind
 and can be placed on different devices. Qwen3-TTS has two stages:
 
 | Stage | Name                | Input           | Output                   |
-|-------|------|-------|--------|
+| ----- | ------------------- | --------------- | ------------------------ |
 | 0     | Code Predictor (AR) | Text tokens     | Discrete RVQ codec codes |
 | 1     | Code2Wav (Decoder)  | RVQ codec codes | Audio waveform           |
 
@@ -170,7 +170,7 @@ and Qwen3-Omni).
 
 When adding a new TTS model, create the following structure:
 
-```
+```text
 vllm_omni/model_executor/models/
   your_model_name/
     __init__.py
@@ -201,7 +201,7 @@ it would not be picked up — the hub README is the documented surface.
 **Qwen3-TTS reference files:**
 
 | File | Purpose |
-|------|---------|
+| ------ | --------- |
 | `models/qwen3_tts/qwen3_tts.py` | Unified model class |
 | `models/qwen3_tts/qwen3_tts_code_predictor_vllm.py` | Stage 0 - optimized AR |
 | `models/qwen3_tts/qwen3_tts_code2wav.py` | Stage 1 - decoder |
@@ -427,112 +427,93 @@ _OMNI_MODELS = {
 
 The registry uses lazy loading - model classes are only imported when needed.
 
-## Stage Configuration
+## Pipeline and Deploy Configuration
 
-Each stage has a `worker_type` that determines how it is scheduled:
+Define fixed topology in `models/<model>/pipeline.py`. The execution type
+selects the scheduler and worker family:
 
-- `worker_type: ar` - autoregressive stage, uses `OmniARScheduler` with PagedAttention
-- `worker_type: generation` - non-AR stage (e.g. decoder), uses `OmniGenerationScheduler`
+- `StageExecutionType.LLM_AR` uses the autoregressive runtime with PagedAttention.
+- `StageExecutionType.LLM_GENERATION` uses the generation runtime for stages such as audio decoders.
 
-Key configuration fields:
+Key topology fields belong to `StagePipelineConfig`:
 
 | Field | Description |
 |-------|-------------|
-| `model_stage` | Which stage to initialize (`ar_stage`, `decoder`, etc.) |
-| `model_arch` | Architecture name, must match `registry.py` |
-| `engine_input_source` | List of upstream stage IDs that provide input (e.g. `[0]`) |
-| `engine_output_type` | Output type: `latent` for intermediate, `audio` for final |
-| `custom_process_next_stage_input_func` | Async chunk processor function path (streaming only) |
-| `final_output` | Whether this stage produces the final user-facing output |
-| `final_output_type` | Type of final output (`audio`, `text`, etc.) |
+| `model_stage` | Logical stage name (`ar_stage`, `decoder`, etc.). |
+| `execution_type` | Runtime family for this stage. |
+| `input_sources` | Upstream stage IDs that provide input. |
+| `model_arch` | Stage-specific architecture name, when it differs from the pipeline default. |
+| `engine_output_type` | Output representation such as `latent` or `audio`. |
+| `custom_process_next_stage_input_func` | Processor for full-payload stage handoff. |
+| `async_chunk_process_next_stage_input_func` | Processor for streaming chunk handoff. |
+| `final_output` / `final_output_type` | Whether this stage produces user-facing output and its modality. |
+| `owns_tokenizer` | Whether this stage supplies the pipeline tokenizer. |
 
-!!! note
-    New in-tree models should define frozen topology in `models/<model>/pipeline.py`,
-    register it in `vllm_omni/config/pipeline_registry.py`, and put deployment
-    knobs in `vllm_omni/deploy/<model>.yaml`. Use `--deploy-config` to load a
-    custom deployment YAML.
+For example, a two-stage TTS topology can be defined as:
 
-### Batch mode
+```python
+from vllm_omni.config.stage_config import (
+    PipelineConfig,
+    StageExecutionType,
+    StagePipelineConfig,
+)
 
-```yaml
-# stage_configs/your_model_name.yaml
+_PROC = "vllm_omni.model_executor.stage_input_processors.your_model_name"
 
-stage_args:
-  - stage_id: 0
-    stage_type: llm
-    runtime:
-      devices: "0"
-    engine_args:
-      model_stage: ar_stage
-      max_num_seqs: 64
-      model_arch: YourTTSModelForConditionalGeneration
-      worker_type: ar
-      scheduler_cls: vllm_omni.core.sched.omni_ar_scheduler.OmniARScheduler
-      engine_output_type: latent
-    default_sampling_params:
-      temperature: 0.9
-      top_k: 50
-      max_tokens: 2048
-
-  - stage_id: 1
-    stage_type: llm
-    runtime:
-      devices: "0"
-    engine_args:
-      model_stage: decoder
-      model_arch: YourTTSModelForConditionalGeneration
-      worker_type: generation
-      scheduler_cls: vllm_omni.core.sched.omni_generation_scheduler.OmniGenerationScheduler
-      engine_output_type: audio
-    engine_input_source: [0]
-    final_output: true
-    final_output_type: audio
+YOUR_TTS_PIPELINE = PipelineConfig(
+    model_type="your_model_name",
+    default_deploy_config_name="your_model_name.yaml",
+    model_arch="YourTTSModelForConditionalGeneration",
+    stages=(
+        StagePipelineConfig(
+            stage_id=0,
+            model_stage="ar_stage",
+            execution_type=StageExecutionType.LLM_AR,
+            input_sources=(),
+            owns_tokenizer=True,
+            engine_output_type="latent",
+            custom_process_next_stage_input_func=f"{_PROC}.ar2decoder",
+            async_chunk_process_next_stage_input_func=f"{_PROC}.ar2decoder_async_chunk",
+        ),
+        StagePipelineConfig(
+            stage_id=1,
+            model_stage="decoder",
+            execution_type=StageExecutionType.LLM_GENERATION,
+            input_sources=(0,),
+            model_arch="YourTTSDecoder",
+            engine_output_type="audio",
+            final_output=True,
+            final_output_type="audio",
+        ),
+    ),
+)
 ```
 
-### Streaming mode (async_chunk)
-
-Add `async_chunk: true` at the top level and specify `custom_process_next_stage_input_func`
-on Stage 0 to define how intermediate outputs are chunked and forwarded:
+Register the pipeline in `vllm_omni/config/pipeline_registry.py`. Runtime
+placement and sizing belong in `vllm_omni/deploy/your_model_name.yaml`:
 
 ```yaml
-# stage_configs/your_model_name_async_chunk.yaml
-
 async_chunk: true
 
-stage_args:
+stages:
   - stage_id: 0
-    stage_type: llm
-    runtime:
     devices: "0"
-    engine_args:
-      model_stage: ar_stage
     max_num_seqs: 64
-      model_arch: YourTTSModelForConditionalGeneration
-      worker_type: ar
-      scheduler_cls: vllm_omni.core.sched.omni_ar_scheduler.OmniARScheduler
-      engine_output_type: latent
-      custom_process_next_stage_input_func: >
-        vllm_omni.model_executor.stage_input_processors.your_model_name.ar2decoder_async_chunk
     default_sampling_params:
       temperature: 0.9
       top_k: 50
       max_tokens: 2048
 
   - stage_id: 1
-    stage_type: llm
-    runtime:
     devices: "0"
-    engine_args:
-      model_stage: decoder
-      model_arch: YourTTSModelForConditionalGeneration
-      worker_type: generation
-      scheduler_cls: vllm_omni.core.sched.omni_generation_scheduler.OmniGenerationScheduler
-      engine_output_type: audio
-    engine_input_source: [0]
-    final_output: true
-    final_output_type: audio
+    max_num_seqs: 1
 ```
 
+Set `async_chunk: false` for end-to-end handoff. The registered pipeline
+selects the full-payload or async chunk processor; processor paths and topology
+must not be moved into the deploy YAML. Use `--deploy-config` to load a custom
+deployment overlay.
+
 ## Stage Input Processors
 
 Stage input processors convert Stage 0 outputs into Stage 1 inputs. Create yours in
@@ -629,6 +610,7 @@ def ar2decoder_async_chunk(
 ```
 
 Key points:
+
 - `transfer_manager` is the `OmniChunkTransferAdapter` that owns the chunk lifecycle
 - Each call appends one AR decode step's output; a chunk is emitted every `chunk_size` steps
 - The final (possibly partial) chunk is flushed when `is_finished` is true
@@ -683,6 +665,7 @@ Create `vllm_omni/entrypoints/openai/tts_adapters/your_model.py`:
 
 ```python
 # SPDX-License-Identifier: Apache-2.0
+# SPDX-FileCopyrightText: Copyright contributors to the vLLM-Omni project
 """YourModel serving adapter."""
 
 from typing import TYPE_CHECKING
@@ -821,7 +804,7 @@ directly from the model's own generator.
 
 ### Directory structure
 
-```
+```text
 vllm_omni/model_executor/models/your_model_name/
     __init__.py
     modeling_your_model_name.py    # unified class: load_weights + forward + streaming
@@ -832,32 +815,40 @@ vllm_omni/deploy/your_model_name.yaml
 
 No stage input processor is needed.
 
-### Stage config
+### Pipeline and deploy config
 
-Use a single stage with `worker_type: ar`. The `is_comprehension: true` field and the
-top-level `async_chunk: false` are required — omitting them causes silent
-misclassification in the serving layer. Set `max_num_seqs` to at least 4 for
-concurrent production use.
+Use a single `LLM_AR` stage that owns the tokenizer and produces final audio:
+
+```python
+YOUR_TTS_PIPELINE = PipelineConfig(
+    model_type="your_model_name",
+    default_deploy_config_name="your_model_name.yaml",
+    model_arch="YourModelForCausalLM",
+    stages=(
+        StagePipelineConfig(
+            stage_id=0,
+            model_stage="your_model_stage_key",
+            execution_type=StageExecutionType.LLM_AR,
+            input_sources=(),
+            owns_tokenizer=True,
+            engine_output_type="audio",
+            final_output=True,
+            final_output_type="audio",
+        ),
+    ),
+)
+```
+
+The deploy YAML controls runtime behavior. Set `max_num_seqs` to at least 4
+for concurrent production use:
 
 ```yaml
-# stage_configs/your_model_name.yaml
 async_chunk: false
 
-stage_args:
+stages:
   - stage_id: 0
-    stage_type: llm
-    is_comprehension: true          # required for serving_speech.py dispatch
-    runtime:
     devices: "0"
-    engine_args:
-      model_stage: your_model_stage_key
-      model_arch: YourModelForCausalLM
-      worker_type: ar
-      scheduler_cls: vllm_omni.core.sched.omni_ar_scheduler.OmniARScheduler
-      engine_output_type: audio
-      max_num_seqs: 4               # min 4 for concurrent requests; default 1 causes gaps
-    final_output: true
-    final_output_type: audio
+    max_num_seqs: 4
 ```
 
 ### Generator-based streaming pattern
@@ -961,10 +952,17 @@ is correct. Stage the modified files and commit again — do not revert the chan
 Common failures and fixes:
 
 | Check | Cause | Fix |
-|-------|-------|-----|
+| ------- | ------- | ----- |
 | `ruff F841` | Local variable assigned but never used | Remove the extraction or forward it to the model call |
 | `ruff E402` | Module-level import not at top of file | Move import to the top-level import block |
 | `ruff format` | Line length, spacing, or quote style | Accept the auto-fix, stage, and re-commit |
+| `check-spdx-header` | Missing header, or copyright still says `vLLM project` | Use `Copyright contributors to the vLLM-Omni project`; restage the rewrite |
+| `check-forbidden-imports` | Stdlib `re`/`base64`, pickle, Hugging Face Hub API, or direct Triton/TileLang | `import regex as re` and `pybase64`; do not grow the allowlist without review |
+| `check-torch-cuda-call` | New `torch.cuda.*` call site | Use `current_omni_platform`; do not grow `ALLOWED_FILES` without review |
+| `check-tts-adapter-migration` | New `_tts_model_type` branch in `serving_speech.py` | Put logic in `tts_adapters/`; lower `MAX_MODEL_TYPE_BRANCHES` when removing branches |
+| `check-test-ci-coverage` | Test file missing level or hardware mark | Add `core_model`/… plus `cpu`/`cuda`/`hardware_test(` |
+| `shellcheck` | Native `shellcheck` missing, or script warning | Install via apt/dnf/brew (or `shellcheck.exe` on Windows). See [Linting](../README.md#linting) |
+| `mypy-3.10` / `markdownlint-cli2` / `check-buildkite` | Types, docs markdown, or Buildkite YAML | See [Linting](../README.md#linting) |
 
 ### DCO sign-off
 
@@ -1019,7 +1017,7 @@ Adding a TTS model to vLLM-Omni involves:
 ### Qwen3-TTS Reference Files
 
 | File | Purpose |
-|------|---------|
+| ------ | --------- |
 | `models/qwen3_tts/qwen3_tts.py` | Unified model class |
 | `models/qwen3_tts/qwen3_tts_code_predictor_vllm.py` | AR stage with vLLM fused ops |
 | `models/qwen3_tts/qwen3_tts_code2wav.py` | Decoder stage with `chunked_decode_streaming()` |
@@ -1031,4 +1029,4 @@ For more information, see:
 
 - [Architecture Overview](../../design/architecture_overview.md)
 - [Async Chunk Design](../../design/feature/async_chunk.md)
-- [Stage Configuration Guide](../../configuration/stage_configs.md)
+- [Pipeline and Deploy Configuration Guide](../../configuration/stage_configs.md)

docs/design/ar_diffusion_pipeline_capability.md

@@ -4,6 +4,9 @@
 > `experimental/ar_diffusion` contract. It follows the ownership and lifecycle
 > direction in [#5137](https://github.com/vllm-project/vllm-omni/issues/5137),
 > but it is not yet part of the stable module-design hierarchy.
+>
+> The end-to-end request, identity, failure, and memory contracts are documented
+> in [Realtime AR-Diffusion sessions](feature/realtime_ar_diffusion.md).
 
 `ARDiffusionEngine` is an opt-in diffusion backend for pipelines that keep
 autoregressive attention KV across requests. The engine only selects
@@ -51,14 +54,13 @@ time, and its inherited batch/step entry points also fail explicitly. Supporting
 those paths requires a batch-aware state-binding contract rather than silently
 bypassing `bind_ar_diffusion_state()`.
 
-The current implementation retains one session per model instance. A request
-for a different `session_id` closes the resident session through the same LRU
-lifecycle used by explicit close and failure cleanup before allocating the new
-one. Self-attention and cross-attention memory are therefore sized for one
-resident session even if the pipeline declares a larger `session_capacity`.
-This prioritizes rollout throughput and prevents block-pool exhaustion before a
-count-based eviction can run. Multi-session residency can be added later with a
-memory-backed capacity calculation or allocation-pressure eviction.
+The implementation may retain multiple sessions per model instance. It selects
+the largest feasible capacity up to the pipeline's `session_capacity`, based on
+self-attention pages, cross-attention allocations, scratch, and declared
+model-owned state. One session is admitted whenever it fits actual free device
+memory; `gpu_memory_fraction` controls additional residency. Requests beyond
+that capacity evict the least-recently-used session through the same lifecycle
+hook used by explicit close and failure cleanup.
 
 Scratch storage is also capability-driven. `frames_per_block` reserves space for
 an uncommitted current video block, while `max_scratch_tokens_per_branch`

docs/design/architecture_overview.md

@@ -335,8 +335,8 @@ The single resolution boundary is implemented by
 `StageConfigFactory.create_from_model()` and
 `VllmOmniConfig.from_pipeline_config()` in
 [`vllm_omni/config`](https://github.com/vllm-project/vllm-omni/tree/main/vllm_omni/config).
-The legacy `stage_args` YAML path remains only for models that have not yet
-migrated to `PipelineConfig` and `DeployConfig`.
+The legacy `stage_args` YAML path has been removed. Model topology now resolves
+through `PipelineConfig`, with runtime overrides supplied by `DeployConfig`.
 
 In the typed path, each stage configuration derives from
 `BaseVllmOmniStageConfig` and is specialized as
@@ -406,7 +406,7 @@ the source for configuration, compatibility, and implementation details.
   stage contract.
 * **Quantization and memory efficiency:** [Quantization](feature/quantization.md)
   resolves per-pipeline or per-component quantization configurations, while
-  [Distributed Layerwise Offload](feature/distributed_layerwise_offload.md)
+  [Distributed Layerwise Offload](feature/offloader/distributed_layerwise_offload.md)
   streams diffusion blocks from host memory within the existing parallel
   topology.
 

docs/design/feature/async_chunk.md

@@ -211,38 +211,39 @@ In sequential mode, each stage must wait for the previous stage to complete enti
 
 ## Configuration
 
-Enable async_chunk in stage configuration YAML:
+Enable async chunk mode in the deploy YAML:
 
 ```yaml
 async_chunk: true
-stage_args:
-  - stage_id: 0
-    engine_args:
-      custom_process_next_stage_input_func: vllm_omni.model_executor.stage_input_processors.qwen3_omni.thinker2talker_async_chunk
-  - stage_id: 1
-    engine_args:
-      custom_process_next_stage_input_func: vllm_omni.model_executor.stage_input_processors.qwen3_omni.talker2code2wav_async_chunk
 ```
 
-### Stage Configuration
+The registered `PipelineConfig` owns the async handoff processor functions.
+For Qwen3-Omni these are `thinker2talker_async_chunk` and
+`talker2code2wav_async_chunk`; deploy YAML only selects the mode and runtime
+sizing.
+
+### Configuration ownership
 
 - `async_chunk: bool`: Enable/disable async chunk mode
-- `custom_process_next_stage_input_func: str`: Path to custom chunk processing function; receives `(transfer_manager, pooling_output, request)`. For qwen3-omni: `thinker2talker_async_chunk`, `talker2code2wav_async_chunk`
-- `stage_connector_config: dict`: Connector configuration
-- `worker_type: str`: Model type, e.g. `"ar"` or `"generation"` (used by OmniChunkTransferAdapter for mode-specific payload handling)
+- `async_chunk_process_next_stage_input_func: str`: Pipeline-owned chunk processor path
+- `execution_type: StageExecutionType`: Pipeline-owned runtime family
+- `connectors: dict`: Deploy-owned connector definitions
 - `max_num_seqs: int`: Maximum number of sequences for concurrent processing in the stage
 
-
 ### Connector Configuration
 
 ```yaml
 connectors:
-  - from_stage: 0
-    to_stage: 1
-    spec:
+  connector_of_shared_memory:
     name: SharedMemoryConnector
-      extra:
-        stage_id: 0
+
+stages:
+  - stage_id: 0
+    output_connectors:
+      to_stage_1: connector_of_shared_memory
+  - stage_id: 1
+    input_connectors:
+      from_stage_0: connector_of_shared_memory
 ```
 
 ### Code2Wav Batch Configuration
@@ -250,12 +251,9 @@ connectors:
 For optimal performance with async_chunk, the code2wav stage should be configured with batching:
 
 ```yaml
-stage_args:
+stages:
   - stage_id: 2  # code2wav stage
-    runtime:
     devices: "1"
-    engine_args:
-      model_stage: code2wav
     max_num_seqs: 64  # Enables batched audio generation
 ```
 

docs/design/feature/async_diffusion_output.md

@@ -117,7 +117,22 @@ After (async D2H):
 
 ### Batch Split
 
-When `execute_model_batch` returns a `COMPUTE_DONE` with a single `async_output_id` for the entire batch, the executor splits it into per-request `async_output_id`s (format `{batch_id}/{request_id}`) via `_batch_split_map`. When `OUTPUT_READY` arrives, the pump extracts per-request results from the batch output and resolves each request's output future independently.
+When `execute_model_batch` returns a `COMPUTE_DONE` with a single `async_output_id` for the entire batch, the executor splits it into per-request `async_output_id`s (format `{batch_id}/{request_id}`) via `_batch_split_map`. When `OUTPUT_READY` arrives, the pump extracts per-request results from the batch output and resolves each request's output future independently. If `OUTPUT_READY` arrives before `execute_batch` registers the split map, `execute_batch` directly adopts the early output.
+
+### Reliability, Lifecycle & Timeout Behavior
+
+1. **Drain Before Memory Release (`drain_async_outputs` / `_ASYNC_OUTPUT_DRAIN_TIMEOUT_S`)**:
+   When workers transition into sleep states (`handle_sleep_task`) or invoke memory-releasing methods, `WorkerProc.drain_async_outputs()` blocks up to `_ASYNC_OUTPUT_DRAIN_TIMEOUT_S` (10.0s) until all in-flight background D2H/SHM packing tasks finish. This acts as a synchronization barrier preventing worker sleep cycles from releasing device tensors while the side CUDA stream is still actively reading them. Draining typically completes within 10–50 ms under normal operation.
+
+2. **Timeout Diagnostics (`_async_output_timeout()`)**:
+   The engine waits for `wait_output_ready()` with a timeout of `_async_output_timeout()` — 600.0s by default, overridable per run with the `VLLM_OMNI_ASYNC_OUTPUT_TIMEOUT` environment variable (a non-numeric or non-positive value is ignored with a warning and the default applies, since this resolves on the request path). If a timeout occurs, the executor logs detailed diagnostic bookkeeping (pending futures, cached outputs, and batch split maps) to identify whether the stall occurred in the worker, pump, or waiter. The packing itself takes milliseconds, but the wait is queued behind the step's GPU work, so its wall-clock duration tracks step time; the bound is generous because a hung engine is surfaced by the worker monitor and `check_health()`, not by this timeout.
+
+3. **Queue Serialization & Atomic Future Resolution**:
+   - `result_mq` writes are serialized with `_result_mq_lock` to protect the single-writer `MessageQueue` against concurrent writes from the worker main loop (`COMPUTE_DONE`) and background thread (`OUTPUT_READY`).
+   - `ResultPumpThread` unpacks shared-memory payloads first, and then atomically resolves waiting futures or populates `_completed_outputs` under `_futures_lock`, avoiding race conditions with concurrent `wait_output_ready()` invocations.
+
+4. **Storage-Aware IPC Packing**:
+   `pack_diffusion_output_shm()` evaluates `max(view_bytes, storage_bytes)` against `_SHM_TENSOR_THRESHOLD` (1 MB) to ensure that per-request slices sharing a large batch storage are moved to POSIX shared memory rather than serialized through pickle over the MessageQueue inline buffer.
 
 ## Model Coverage
 
@@ -131,6 +146,7 @@ All models running in request-mode (`step_execution=False`, the default) automat
 |---|---|---|
 | **HunyuanImage-3.0** | Image | `False` |
 | **Qwen-Image** | Image | `True` |
+| **LTX-2.3** | Video / Audio | `False` |
 
 Other models with `step_execution=False` are also supported but not yet verified.
 

docs/design/feature/attention_backend_selection.md

@@ -0,0 +1,106 @@
+# Diffusion Attention Backend Selection
+
+This document defines the selection and extension contract for diffusion
+attention backends. User-facing backend choices, installation, and tuning are
+in the [attention backend guides](../../user_guide/diffusion/attention_backends.md).
+
+## Scope
+
+The contract applies to DiT and other diffusion attention layers. It is
+separate from vLLM's autoregressive attention selector.
+
+The selection path has four responsibilities:
+
+1. normalize user configuration into `AttentionConfig` and `AttentionSpec`;
+2. resolve a spec for an attention role;
+3. ask the active platform to validate an explicit backend or choose a
+   hardware default; and
+4. load the selected `AttentionBackend` class from the registry path.
+
+## Resolution contract
+
+`get_attn_backend_for_role()` resolves configuration in this order:
+
+1. exact `per_role[role]`;
+2. category `per_role[role_category]`;
+3. `default`;
+4. platform default.
+
+An explicit resolution returns both the backend class and its
+`AttentionSpec`. A platform-default resolution returns the class and `None`.
+Layers must therefore treat the spec as optional and must not infer that a
+missing spec means SDPA.
+
+Class resolution is cached by backend name, head size, and whether TRTLLM may
+be selected automatically. Logging is cached separately by role and source so
+the startup record identifies why each role received its backend.
+
+## Registry and platform boundary
+
+`DiffusionAttentionBackendEnum` maps stable configuration names to default
+qualified class paths. `register_diffusion_backend()` may replace a path at
+runtime without changing the public enum value.
+
+The active `OmniPlatform` owns compatibility policy through
+`get_diffusion_attn_backend_cls()`. It must:
+
+- validate explicit selections and fail with an actionable error when the
+  requested kernel cannot run;
+- choose only an available, compatible default when no backend is explicit;
+- account for head size and other platform-visible constraints; and
+- return a qualified class path implementing `AttentionBackend`.
+
+The selector must not duplicate device capability or package-availability
+policy that belongs to the platform.
+
+## Typed backend options
+
+Backend-specific settings remain typed fields on `AttentionSpec` rather than
+unstructured keyword dictionaries:
+
+- `quant` is consumed by FlashInfer and TRTLLM with backend-specific value
+  validation;
+- `skip_softmax` is consumed by TRTLLM; and
+- `block_sparse` is consumed by block-sparse backends such as RainFusion; and
+- `fastvideo_vsa_topk` is consumed by FastVideo VSA.
+
+A backend reads only the fields it owns and rejects incompatible values. New
+options should be added to a shared typed spec only when more than one backend
+shares their semantics; otherwise add a dedicated typed block.
+
+## Model integration contract
+
+Each diffusion attention call declares a stable role and, when appropriate, a
+broader category. Model-specific roles permit precise overrides; categories
+preserve common `self` or `cross` policies. A model also decides whether its
+path is compatible with automatic TRTLLM selection, because only the model
+knows whether masking and packing satisfy the kernel contract.
+
+## Adding or changing a backend
+
+An implementation change is complete only when it:
+
+1. implements the `AttentionBackend` interface and exposes a stable enum name;
+2. adds platform validation and default routing where appropriate;
+3. defines and validates any typed configuration;
+4. tests explicit selection, default selection, and incompatible paths;
+5. documents installation, fallback behavior, and quality implications in the
+   corresponding user guide; and
+6. updates the overview matrix without moving algorithm details back into the
+   landing page.
+
+## FastVideo VSA model metadata
+
+FastVideo VSA support is scoped to `FastVideo/FastWan2.2-TI2V-5B-Diffusers`, whose text-to-video and image-to-video modes both use `Wan22Pipeline`. It operates on a flattened DiT sequence but partitions tokens in
+the original latent video grid. Wan integrations therefore attach the
+post-patch `(T, H, W)` shape as `vsa_dit_seq_shape` attention metadata. The
+backend validates that `T * H * W` equals the sequence length, derives the
+runtime block count, and selects the configured top-k key/value blocks for
+each query block.
+
+Checkpoint-specific compensation remains model-owned. Wan layers pass an
+optional learned `gate_compress` projection through attention metadata when
+the checkpoint contains those weights. Checkpoints without the projection do
+not allocate or execute it. When top-k selects every block, native checkpoints
+route to SDPA, while FastVideo DMD checkpoints preserve the VSA all-block path
+to retain checkpoint semantics.

docs/design/feature/disaggregated_inference.md

@@ -59,11 +59,10 @@ This `metadata` must be passed through the control plane so `get()` can locate t
 
 ## Configuration Model
 
-Define connectors in runtime:
+Define connectors at the top level of the deploy YAML:
 
 ```yaml
-runtime:
-  connectors:
+connectors:
   connector_of_shared_memory:
     name: SharedMemoryConnector
 ```
@@ -71,7 +70,7 @@ runtime:
 Wire stages to connectors:
 
 ```yaml
-stage_args:
+stages:
   - stage_id: 0
     output_connectors:
       to_stage_1: connector_of_shared_memory

docs/design/feature/distributed_layerwise_offload.md

@@ -1,156 +0,0 @@
-# Distributed Layerwise Offload
-
-This document describes distributed layerwise offload (DLO) for diffusion
-models. DLO keeps only a small number of DiT blocks on the accelerator and
-streams the remaining blocks from host memory. The distributed backend can
-either shard those host-side weights across an existing parallel group or keep
-the standard loader's rank-local weights and avoid an additional collective.
-
-For user-facing commands, see the
-[distributed layerwise offloading guide](../../user_guide/diffusion/cpu_offload.md)
-and the [Cosmos3 recipe](https://github.com/vllm-project/vllm-omni/blob/main/recipes/cosmos3/Cosmos3-DistOffload.md).
-
-## Status
-
-DLO is implemented for multi-device diffusion execution. The default
-AllGather path is the primary path for DP and SP deployments. The
-`--dlo-no-use-allgather` path is a rank-local compatibility mode: it is useful
-for standard-loader sharding, workstation bring-up, and systems where an
-additional DLO collective is undesirable, but it does not reduce host weight
-storage across ranks.
-
-The compatibility matrix below describes the current implementation. The
-unit-level guards are covered, but not every parallelism combination has a
-full model-and-hardware end-to-end test.
-
-## Design
-
-### DLO consumes the existing parallel topology
-
-DLO does not create a new DP, TP, or SP topology. It reads the configured
-`DiffusionParallelConfig` and attaches offload hooks to the DiT blocks after the
-standard distributed groups have been initialized.
-
-The DLO weight-sharding group is selected as follows:
-
-1. Use the existing DP group when `data_parallel_size > 1`.
-2. When DP is one and SP is greater than one, use the SP group.
-3. Otherwise, run rank-locally without a DLO process group.
-
-TP is deliberately not used as DLO's AllGather group. HSDP has its own
-parameter-sharding lifecycle and is not allowed to be sharded a second time by
-DLO's AllGather path.
-
-### AllGather path
-
-With the default `dlo_use_allgather=True`, each rank stores approximately
-`1 / group_size` of each streamable block in pinned host memory. The next
-block's shard is copied to a device buffer and reconstructed with
-`all_gather_into_tensor` on a communication stream while the current block is
-executing.
-
-```text
-Compute:    [Block N]             [Block N+1]          [Block N+2]
-H2D:                      [shard N+1]           [shard N+2]
-AllGather:                [full N+1]             [full N+2]
-Buffers:    [current slot]       [prefetch slot]       [current slot]
-```
-
-![DLO double-buffer prefetch pipeline](../figures/dlo/dlo_pipeline.gif)
-
-The backend uses two shared device buffers, so accelerator weight residency is
-bounded by the largest streamed blocks rather than the complete model.
-
-When DP is greater than one, the engine can process one request per DP rank in
-the same denoising wave. Because AllGather is a collective, all participating
-requests must take the same execution path at every denoising step.
-
-### Rank-local path without DLO AllGather
-
-With `--dlo-no-use-allgather`, DLO forces its internal offload shard size to
-one. The regular model loader remains responsible for preparing each rank's
-weights, including TP-local tensors or HSDP-managed parameters. DLO then
-streams those rank-local tensors block by block using H2D copies only.
-
-This mode means:
-
-- DP still provides independent replicas, but DLO does not shard weights
-  across DP ranks.
-- SP still performs its normal activation/attention collectives, but DLO does
-  not shard weights across SP ranks.
-- TP/HSDP/SP collectives, if configured, are not disabled by this flag; only
-  DLO's additional weight AllGather is disabled.
-- Pure DP deployments keep a full host-side model copy per rank, subject to
-  shared OS page-cache behavior.
-- The scheduler does not require a synchronized DP request wave for DLO.
-
-This path is intentionally not implemented by reusing the AllGather mmap
-loader. It relies on the standard loader so model-specific TP/HSDP transforms
-remain intact.
-
-## Parallelism compatibility
-
-| Parallelism | DLO + AllGather | DLO without AllGather |
-|---|---|---|
-| **DP** | Supported primary path. DLO shards host weights across the DP group and can run DP multi-concurrency. | Supported rank-local path. DP replicas remain independent; no DLO weight sharding or cross-DP DLO collective. |
-| **SP** | Supported in the implementation. With DP=1, DLO uses the SP group for host-weight sharding; SP still shards sequence/activation work. | SP remains active, but DLO keeps standard-loader rank-local weights and adds no SP weight collective. |
-| **TP > 1** | Unsupported in the DLO mmap path. TP-aware loader callbacks are bypassed, so the backend rejects this configuration when it enters that path. | Standard loading is retained and TP-local tensors can be streamed. This is the intended compatibility path, but it still needs broader model and hardware validation. |
-| **HSDP** | Rejected. HSDP has already sharded parameters, so DLO AllGather would double-shard them. | Accepted by configuration. HSDP owns parameter sharding and its own gathers; DLO only stages rank-local parameters. End-to-end coverage is limited. |
-
-### Combined dimensions
-
-- **DP + SP:** DLO uses the DP group for weight sharding when DP is greater
-  than one; SP continues to use its own sequence-parallel group. If DP is one,
-  the SP group becomes DLO's sharding group in AllGather mode.
-- **DP + TP/SP without AllGather:** standard model loading defines the
-  rank-local tensor layout. DLO adds no cross-DP, cross-TP, or cross-SP weight
-  collective.
-- **HSDP + SP:** the general parallel configuration permits HSDP over SP, but
-  DLO must use `--dlo-no-use-allgather`. HSDP remains responsible for weight
-  materialization and synchronization.
-- **HSDP + DP or TP:** rejected independently by the diffusion parallel
-  configuration.
-
-## Request and loading constraints
-
-AllGather DP multi-concurrency requires:
-
-- explicit `num_inference_steps`;
-- the same `num_inference_steps` for all requests in a wave; and
-- identical request arguments that affect the collective execution path.
-
-The no-AllGather path does not impose these DLO-specific synchronized-wave
-requirements.
-
-The mmap loader is used only by the supported DLO+AllGather path when the
-model has a compatible checkpoint layout. Online quantization is incompatible
-with that sharded mmap path; use `--dlo-no-use-allgather` or disable online
-quantization. Models that do not meet the mmap requirements use the regular
-loader path.
-
-## Validation coverage
-
-Current source-level validation includes:
-
-- HSDP + DLO + AllGather rejection;
-- HSDP + DLO without AllGather acceptance at configuration level;
-- TP rejection in the DLO+AllGather mmap path;
-- resident-layer requests requiring no-AllGather;
-- DP request-wave validation for denoising-step compatibility;
-- sharding, double-buffer, AllGather-size, and heterogeneous-block regression
-  tests.
-
-The highest-value missing coverage is end-to-end numerical comparison against
-ordinary layerwise offload for DP+SP, TP+no-AllGather, and HSDP+SP+no-AllGather
-on the target CUDA/NCCL or CANN/HCCL hardware.
-
-## Recommendations
-
-- Use **DP + DLO AllGather** for the supported throughput and host-memory
-  scaling path.
-- Use **SP + DLO AllGather** for long-sequence workloads when DP concurrency is
-  not the goal.
-- Use **no-AllGather** to bring up TP or workstation PCIe configurations, with
-  the expectation of higher host-memory use and lower validation confidence.
-- Prefer **HSDP alone** for production HSDP deployments until the combined
-  HSDP + DLO no-AllGather path has broader end-to-end coverage.

docs/design/feature/host_weight_runtime.md

@@ -0,0 +1,386 @@
+# Host Weight Runtime
+
+The Host Weight Runtime is a loader-facing feature for reusing immutable,
+runtime-ready model weights across workers on one host. It lets a model
+integration replace repeated checkpoint loading and transformation with an
+exact lookup of a previously published host representation, while keeping the
+canonical model source authoritative.
+
+This feature is shared infrastructure. Diffusion, autoregressive, speech, and
+future model integrations may depend on the top-level
+`vllm_omni.host_weight_runtime` package without depending on one another's
+loader or execution stack. The internal component and filesystem contracts are
+specified in the [Host Weight Runtime module design](../module/host_weight_runtime.md).
+
+## Status
+
+The first implementation provides contracts and a CPU local-filesystem store.
+The initial diffusion consumer contract additionally defines typed,
+representation-independent final-layout identity/restoration mechanics plus a
+concrete BF16-with-preserved-FP32 policy for MiniMax H3. It remains
+library-only: no loader or DLO path selects, publishes, restores, or transports
+that artifact yet.
+
+V1 includes:
+
+- exact runtime-weight identity and immutable manifests;
+- coordinated local lookup and one-producer publication;
+- descriptor-backed safetensors mmap leases;
+- preferred and required resolution policy;
+- explicit, separately reported post-load publication;
+- validation, deny, quarantine, cleanup, and capacity controls; and
+- typed reports for every terminal resolution outcome.
+
+V1 does not include:
+
+- a public CLI, loader activation, or default-on model integration;
+- lease handoff to DLO or another transport consumer;
+- FP8, quantized, merged-adaptation, or additional model producers;
+- CUDA registration, pinned staging, H2D scheduling, or GPU kernels;
+- a remote artifact provider or cross-node coordination;
+- automatic eviction; or
+- a change to DLO AllGather or no-AllGather behavior.
+
+## Motivation and use cases
+
+Model loading can create the same final host representation repeatedly. This
+is especially expensive when loading performs checkpoint decoding, tensor
+renaming, TP slicing, quantization, packing, or scale construction before GPU
+transfer. Independent workers may then retain private CPU copies of identical
+runtime weights.
+
+The Host Weight Runtime is useful when:
+
+- multiple same-node workers request semantically identical final weights;
+- constructing those weights is materially more expensive than validating and
+  mapping a local artifact;
+- the runtime layout can be identified exactly and reproduced deterministically;
+- the workers can share immutable file-backed pages through the OS page cache;
+  and
+- canonical loading remains available when policy permits fallback.
+
+Typical consumers include same-node diffusion replicas, DLO host-weight
+sources, and future transformed or quantized model loaders. Request routing and
+replica orchestration remain outside this feature.
+
+This feature is not zero-copy GPU execution. A lease exposes stable CPU tensor
+views and mapped ranges. A separate transport still decides whether to use
+registered mmap, private pinned staging, synchronous copies, or asynchronous
+H2D transfer.
+
+## Resolution behavior
+
+The loader resolves the immutable canonical source and computes the exact
+requested identity before asking the runtime for a representation.
+
+```mermaid
+flowchart TD
+    L["Loader resolves canonical source"] --> I["Compute exact WeightArtifactIdentity"]
+    I --> M{"Runtime mode"}
+    M -->|"disabled"| C["Canonical loader"]
+    M -->|"preferred or required"| A{"Validated local artifact?"}
+    A -->|"yes"| R["Return HostWeightLease"]
+    A -->|"no"| X["Remote exact artifact (future)"]
+    X --> P{"Registered producer available?"}
+    P -->|"yes"| B["Build, validate, and publish atomically"]
+    B --> R
+    P -->|"no or recoverable failure"| F{"Policy permits fallback?"}
+    F -->|"preferred"| C
+    F -->|"required"| E["Fail startup"]
+    P -->|"nonretryable failure"| E
+    C --> W{"Post-load publication enabled?"}
+    W -->|"yes"| B2["Publish final model through POST_LOAD_ONLY producer"]
+    B2 --> CL["Close validated publication lease"]
+    CL --> R2["Return separate publication report"]
+    R2 --> D
+    W -->|"no"| D["Keep canonical model"]
+```
+
+The modes are:
+
+| Mode | Observable behavior |
+| --- | --- |
+| `disabled` | Use canonical loading directly without probing storage, identity, credentials, or topology. |
+| `preferred` | Return an exact lease when available; otherwise use canonical fallback only for a miss, invalid cache entry, or typed retryable failure. |
+| `required` | Fail when an exact lease cannot be acquired. |
+
+During pre-load resolution, an unsupported capability, semantic identity
+collision, producer failure, or publication failure is nonretryable and remains
+visible even in preferred mode. A post-load publication failure is likewise
+visible in its own report, but cannot revise the canonical-fallback outcome.
+Storage policy must never disguise a semantic or configuration error as a cache
+miss.
+
+## Loader and restoration sequence
+
+The model integration owns the end-to-end loading transaction. The runtime
+does not construct a model or invoke the canonical loader.
+
+```mermaid
+sequenceDiagram
+    participant L as Loader adapter
+    participant R as HostWeightRuntime
+    participant S as HostWeightStore
+    participant P as WeightProducer
+    participant X as WeightRestorer
+    participant T as GPU transport
+
+    L->>L: Resolve canonical revision and exact identity
+    L->>R: resolve(identity, producer)
+    R->>S: lookup(exact identity)
+    alt validated hit
+        S-->>R: HostWeightLease
+        R-->>L: lease and LOCAL_HIT report
+        L->>X: plan_restore(model, lease)
+        X-->>L: validation-only restore plan
+        L->>X: commit() once
+    else miss with allowed producer
+        R->>S: get_or_build(identity, producer)
+        S->>P: produce(store-scoped writer)
+        P-->>S: final-layout tensors and metadata
+        S-->>R: validated HostWeightLease
+        R-->>L: lease and LOCAL_PRODUCTION report
+        L->>X: plan_restore(model, lease)
+        X-->>L: validation-only restore plan
+        L->>X: commit() once
+    else policy permits canonical fallback
+        R-->>L: CANONICAL_FALLBACK
+        L->>L: Run canonical loader
+        opt explicit post-load publication enabled
+            L->>R: publish_after_load(identity, POST_LOAD_ONLY producer)
+            R->>S: get_or_build(identity, producer)
+            S-->>R: validated HostWeightLease or typed failure
+            R->>R: close publication lease
+            R-->>L: separate publication report
+        end
+    else required or nonretryable failure
+        R-->>L: FAILED report
+    end
+    L->>T: consume final CPU tensors and any lease
+    T-->>L: transfer teardown complete
+    L->>L: close lease when one was acquired
+```
+
+`plan_restore()` must not mutate the model or lease. `commit() -> None` is the
+sole one-shot model mutation. If planning fails, canonical fallback may reuse
+the untouched model. If commit begins and fails, the partially hydrated model
+must be discarded and canonical fallback must construct a fresh model.
+
+`publish_after_load()` is synchronous in V1 and accepts only a
+`POST_LOAD_ONLY` producer. Its report is separate from the terminal resolution,
+so a publication failure cannot rewrite a successful canonical fallback. A
+successful publication closes the store-returned lease inside the runtime and
+warms only future startups. It does not restore, rebind, or otherwise mutate the
+canonically loaded model serving the current startup.
+`allow_local_build` gates producers during pre-load resolution, while
+`allow_post_load_publish` independently gates this explicit post-load path.
+
+## Exact representation identity
+
+The store performs exact lookup. It never silently converts one representation
+into another or substitutes the first backing that responds.
+
+Identity includes:
+
+- immutable model revision and source fingerprint;
+- model component and ownership boundary;
+- representation name, dtype, and format metadata;
+- final tensor layout and semantic parallel coordinates;
+- static adaptation identity; and
+- producer implementation, manifest, and restorer schema versions.
+
+The requested representation is selected by the loader. The store chooses
+where to obtain that exact representation: validated local artifact, future
+remote materialization, or a registered producer.
+
+### Parallelism
+
+Parallel coordinates are included only when they change bytes, shape, layout,
+or component ownership.
+
+| Dimension | Identity and sharing rule |
+| --- | --- |
+| DP | Exclude DP rank for replicated weights so same-node replicas can acquire the same artifact. Include a coordinate only if DP changes weight ownership. |
+| TP | Include TP size and rank when each rank owns different tensor slices or packed layouts. Each distinct TP shard is a separate exact artifact. |
+| SP | Include SP size and backend when they change the weight layout. Exclude SP rank when all SP ranks consume identical bytes. |
+| PP | Encode component ownership or stage-local layout when PP changes which weights belong to the consumer. |
+| EP | Encode expert ownership and layout whenever ranks own different expert weights. |
+
+The runtime does not infer these rules from process-group topology. The model
+adapter is responsible for constructing a correct identity.
+
+### Quantization and adaptations
+
+Runtime BF16, FP8, packed quantized weights, and future formats are distinct
+representations. Conversion is performed by an explicitly versioned producer,
+never by cache coercion. A producer must publish the final layout consumed by
+its matching restorer.
+
+Dynamic LoRA overlays are not part of a reusable base-weight artifact. A
+statically merged adapter is cacheable only as a separate identity containing
+the adapter fingerprint and merge semantics.
+
+### Initial diffusion final-layout contract
+
+The shared diffusion contract covers complete final-layout DiT parameters and
+persistent buffers. Text encoders, VAEs, non-persistent derived state, and other
+pipeline components remain outside the artifact. One explicit representation
+policy selects allowed dtypes, tensor roles, physical layout identity, producer
+ABI, manifest schema, and restoration schema.
+
+The contract is intentionally separate from loader activation:
+
+- `FinalLayoutRequest` contains typed loader identity/configuration fingerprints,
+  TP coordinate, and conservative SP semantics. It has no open metadata bag,
+  DP coordinate, SP rank, device identity, DLO transfer mode, registration
+  policy, or store path.
+- `FinalLayoutArtifactSpec` binds one `WeightRepresentation` and runtime-layout
+  name to explicit producer/restorer schemas and a canonical, versioned
+  implementation ABI descriptor. Compatibility never depends on reflective
+  source inspection.
+- `PreparedWeightSource` snapshots immutable revisions or exact local file
+  content plus a typed checkpoint-adapter identity before ordinary
+  materialization. Source replacement before or during production fails
+  publication. A hash-looking symlink basename is trusted only for an explicit
+  Hugging Face Hub source whose repository ID and
+  `models--.../snapshots/<revision> -> blobs/<hash>` topology validate; every
+  local or otherwise unverified symlink target is content-hashed.
+- the tensor ownership digest records exact runtime names, kinds, shapes,
+  semantic roles, dtypes, and strides from a CPU or meta model skeleton;
+- `FinalLayoutTensorRestorer` accepts only an exact lease identity, validates
+  complete policy-defined coverage without mutation, and returns a one-shot
+  commit plan;
+- each model declares one dtype-neutral `FinalLayoutModelContract` with an
+  explicit implementation version and a post-commit validator; and
+- `FinalLayoutBF16Producer` accepts only the matching identity context and a
+  finalized CPU model. It is `POST_LOAD_ONLY` and `SINGLE_PROCESS` per exact TP
+  coordinate. Its BF16 policy preserves model-declared FP32 parameters and
+  buffers and revalidates MiniMax H3 mixed-precision invariants.
+
+Other representations reuse source identity, typed parallel identity, tensor
+ownership, and exact restoration only when their policy proves those semantics.
+For example, runtime FP8 needs a separate policy/producer for generated scales,
+quantization metadata, and Cutlass physical layouts; it is not enabled by
+changing a dtype string on the BF16 producer.
+
+This stage makes no startup, sharing, or DLO performance claim. A following
+consumer PR owns disabled/preferred/required precedence, mixed-component loader
+transactions, warm-hit restoration, and transactional lease handoff. A TP2
+prewarm deployment will require a matching TP2 producer cohort to populate both
+TP-coordinate identities even though the store coordinates each artifact
+independently.
+
+## Host sharing and GPU transport
+
+Every process receives its own virtual mappings and `HostWeightLease`, but
+workers mapping the same local artifact can share physical file-backed pages
+through the OS page cache. This avoids independent private tensor allocations;
+it does not guarantee that process PSS reports exactly one model copy.
+
+The ownership boundary is:
+
+- the store owns immutable artifact files, publication, and lifecycle;
+- the kernel owns page-cache residency and placement;
+- the lease owns process-local mappings and the shared artifact lock; and
+- transport owns page registration, page locking, private staging, H2D copies,
+  streams, device buffers, and lease release ordering.
+
+DLO AllGather and no-AllGather remain transport and execution choices. A DLO
+integration may consume a lease, but the Host Weight Runtime does not choose a
+parallel collective or orchestrate DP requests. See the
+[DLO feature design](offloader/distributed_layerwise_offload.md).
+
+## Locality and NUMA policy
+
+The V1 backend is node-local. It accepts an allowlist of known local disk
+filesystems plus tmpfs/ramfs, records the detected kernel filesystem type, and
+rejects known remote or unknown filesystems. NFS, CIFS, Lustre, Ceph, and other
+cross-node mounts cannot silently satisfy the local backend because their
+page-cache and advisory-lock behavior does not provide the required node-local
+contract.
+
+A NUMA-specific root creates a separate storage domain but does not by itself
+place page-cache pages on that NUMA node. A topology-aware integration must
+also pin workers, establish local first touch or prefault, perform transport
+registration in the same domain, and collect residency evidence.
+
+Tmpfs artifacts consume host memory and may consume swap. They must be
+accounted as memory-backed storage rather than ordinary disk capacity.
+
+## Concurrency and timeouts
+
+One process per exact identity owns a build; other workers wait and then acquire
+leases for the published artifact. Publication is invisible until all payloads
+and metadata are validated, hashed, fsynced, and atomically renamed.
+
+`coordination_timeout_seconds` bounds filesystem lock acquisition. It does not
+cancel synchronous validation, a producer that has already started, or atomic
+publication. A hung in-process producer therefore blocks its owning process and
+must be handled by external process supervision. Enforceable producer
+cancellation requires a future process-isolated producer contract.
+
+Leases keep a shared artifact lock for their lifetime. A forked child may close
+its inherited descriptors but cannot unlock the parent process's lease. Cleanup
+uses a nonblocking exclusive lock and reports an active lease instead of
+removing live mappings.
+
+## Failure and fallback contract
+
+| Condition | Preferred mode | Required mode |
+| --- | --- | --- |
+| Exact local hit | Return lease | Return lease |
+| Local miss | Try later backing or canonical fallback | Try later backing or fail |
+| Invalid or denied artifact | Quarantine/rebuild when possible; otherwise canonical fallback | Quarantine/rebuild or fail |
+| Retryable lock, domain, or capacity failure | Canonical fallback | Fail |
+| Unsupported producer or backend | Fail | Fail |
+| Identity collision | Fail | Fail |
+| Pre-load producer or atomic publication failure | Fail | Fail |
+| Post-load publication failure after canonical fallback | Keep canonical model; report publication failure | Not reached through required resolution |
+| Restore planning failure | Canonical fallback may reuse untouched model | Fail |
+| Restore commit failure | Discard model; fallback requires a fresh instance | Fail |
+
+Cache failures should normally reduce performance rather than availability,
+but only when the typed failure explicitly permits fallback. Canonical source,
+authentication, configuration, and semantic errors remain visible.
+
+## Consumer integration requirements
+
+A consumer PR must:
+
+1. Resolve an immutable canonical model source before cache lookup.
+2. Construct the exact final representation identity, including relevant
+   parallel, quantization, and adaptation semantics.
+3. Register a deterministic producer for that identity or support lookup-only
+   operation.
+4. Provide a validation-only restorer with a one-shot commit.
+5. Keep GPU transport outside the producer, store, and restorer contracts.
+6. Handle preferred fallback and required failure without reusing a model after
+   a failed restore commit.
+7. Retain the lease until every transport operation that may access mapped
+   memory has completed.
+8. Emit and test the terminal resolution report and any separate post-load
+   publication report.
+
+Consumer validation must prove:
+
+- output parity with canonical loading;
+- exact warm-hit work avoidance;
+- correct TP/SP/PP/EP ownership and layout identity;
+- correct BF16, FP8, quantized, or adaptation semantics;
+- shared-backing evidence when host-memory savings are claimed;
+- clean unmapping, unregistration, and artifact-lock release; and
+- startup, host-memory, and transport effects for the claimed deployment.
+
+## Rollout
+
+The rollout is intentionally staged:
+
+1. Land the neutral contracts and local filesystem store.
+2. Add one model-specific producer/restorer integration with parity evidence.
+3. Connect eligible DLO or other transport paths without moving transport
+   ownership into the runtime.
+4. Add additional final-layout and quantized producers independently.
+5. Introduce remote materialization only through the same local lease contract.
+
+The architectural decisions and deferred work are tracked in
+[RFC #6414](https://github.com/vllm-project/vllm-omni/issues/6414).

docs/design/feature/offloader/README.md

@@ -0,0 +1,80 @@
+# CPU Offloading
+
+This document defines the shared architecture for diffusion CPU offloading.
+The three strategies have separate user and design pages:
+
+| Strategy | User guide | Design contract |
+| --- | --- | --- |
+| Model-level | [Guide](../../../user_guide/diffusion/offloader/cpu_offload.md) | [Design](cpu_offload.md) |
+| Layerwise | [Guide](../../../user_guide/diffusion/offloader/layerwise_offload.md) | [Design](layerwise_offload.md) |
+| Distributed layerwise | [Guide](../../../user_guide/diffusion/offloader/distributed_layerwise_offload.md) | [Design](distributed_layerwise_offload.md) |
+
+## Strategy selection
+
+`OffloadConfig.from_od_config()` converts diffusion configuration into one
+`OffloadStrategy`. If callers set multiple flags, selection is deterministic:
+
+1. `DISTRIBUTED_LAYERWISE`;
+2. `LAYERWISE`;
+3. `MODEL_LEVEL`; and
+4. `NONE`.
+
+The factory logs when a higher-priority strategy disables another. It derives
+parallel and HSDP state from `DiffusionParallelConfig`; callers do not provide
+a separate offload group size.
+
+`get_offload_backend()` then validates platform offload support, resolves the
+device, and creates exactly one backend. Returning `None` means offloading is
+disabled or unsupported and must not leave partially installed hooks.
+
+## Shared lifecycle
+
+Every backend implements `OffloadBackend`:
+
+- `enable(pipeline)` discovers modules, establishes initial residency, and
+  installs hooks;
+- `disable()` removes owned hooks and resources; and
+- `is_enabled()` reports lifecycle state.
+
+`disable()` does not promise to restore the pipeline's original device
+placement. The caller owns any subsequent rematerialization.
+
+Hooks are registered through `HookRegistry` and `ModelHook`; offload backends
+must use distinct hook names and remove only hooks they own.
+
+## Discovery and topology
+
+Pipeline component discovery prefers `SupportsComponentDiscovery` declarations:
+
+- `_dit_modules`;
+- `_encoder_modules`;
+- `_vae_modules`; and
+- `_resident_modules`.
+
+Dotted paths are supported. Legacy pipelines may use the fallback scan of
+well-known attribute names, but new integrations should declare components
+explicitly.
+
+Ordinary layerwise offload obtains topology from declared
+`_layerwise_offload_blocks_attrs` (with a deprecated singular-name
+compatibility path). Distributed layerwise offload may instead use a pipeline
+`OffloadPlan`, then falls back to declared block attributes and its documented
+compatibility discovery. Discovery metadata describes structure only; the
+backend remains responsible for transfer, synchronization, and storage
+ownership.
+
+## Cross-strategy invariants
+
+- At most one strategy owns offload hooks for a pipeline.
+- A parameter has one authoritative host representation while offloaded.
+- Device storage is not freed until dependent compute or transfer work is
+  complete.
+- Non-persistent and model-specific buffers remain correct after movement or
+  checkpoint rematerialization.
+- Unsupported parallel or loading combinations fail before hooks mutate the
+  model.
+- Platform streams, events, synchronization, and cache management go through
+  the vLLM-Omni platform abstraction.
+
+The diffusion [Offloader module design](../../module/diffusion/offloader.md)
+describes how these feature contracts fit into the larger diffusion runtime.

docs/design/feature/offloader/cpu_offload.md

@@ -0,0 +1,56 @@
+# Model-Level Offload
+
+Model-level offload enforces mutual exclusion between pipeline component
+groups. For commands and model support, see the
+[model-level user guide](../../../user_guide/diffusion/offloader/cpu_offload.md).
+
+## Generic hook path
+
+`ModelLevelOffloadBackend.enable()` discovers DiTs, encoders, and VAEs. It
+makes encoders and VAEs device resident for initialization, then installs
+`SequentialOffloadHook` instances:
+
+- each DiT hook offloads all encoders and other DiTs before loading itself;
+- each encoder hook offloads every DiT before loading itself; and
+- VAEs are not part of the mutual-exclusion hook set.
+
+The hook moves parameters and buffers without recursively calling
+`module.to()`. This avoids recursion through modules that retain references to
+other transformer blocks. Host tensors are pinned when configured; DTensors
+are not pinned through the ordinary tensor path.
+
+Before forward returns control to the model, the hook synchronizes the active
+platform so compute cannot race the transfer. XPU uses blocking device-to-host
+copies because its allocator cannot safely evict storage while an asynchronous
+copy is pending.
+
+## Split-model delegation
+
+If a pipeline provides callable `enable_omni_model_cpu_offload`, the backend
+delegates instead of installing generic encoder/DiT hooks. This supports models
+whose mutually exclusive components are nested inside one transformer, such
+as a reasoner and generator.
+
+The delegated pipeline owns its internal contexts, but must preserve the same
+contract:
+
+- only the executing component is device resident;
+- enable and disable are idempotent;
+- transfers use the supplied device, pinning, and HSDP settings; and
+- disabling removes all model-local offload hooks or contexts.
+
+The delegation is currently duck typed. If another independent implementation
+is added, formalize the enable/disable pair as a protocol rather than expanding
+attribute-name heuristics.
+
+## Extension invariants
+
+- New ordinary pipelines declare both DiT and encoder groups through
+  `SupportsComponentDiscovery`.
+- Models with multiple DiTs offload the inactive DiTs as well as encoders.
+- Empty or parameterless modules do not trigger transfers.
+- Hook removal does not move modules; callers must not assume original
+  residency is restored.
+
+Shared selection and lifecycle behavior is defined in the
+[CPU Offloading design](README.md).

docs/design/feature/offloader/distributed_layerwise_offload.md

@@ -0,0 +1,494 @@
+# Distributed Layerwise Offload
+
+This document describes distributed layerwise offload (DLO) for diffusion
+models. DLO keeps only a small number of DiT blocks on the accelerator and
+streams the remaining blocks from host memory. The distributed backend can
+either shard those host-side weights across an existing parallel group or keep
+complete rank-local block sources and avoid an additional collective.
+
+For user-facing commands, see the
+[distributed layerwise offloading guide](../../../user_guide/diffusion/offloader/distributed_layerwise_offload.md)
+and the [Cosmos3 recipe](https://github.com/vllm-project/vllm-omni/blob/main/recipes/cosmos3/Cosmos3-DistOffload.md).
+
+## Feature compatibility
+
+Host-storage optimization and runtime compatibility are separate decisions.
+When direct checkpoint mmap is unavailable, DLO can still use tensors produced
+by the ordinary loader. "Compatibility path" below means that fallback is
+implemented but has less end-to-end coverage than the primary path.
+
+Legend: ✅ supported, ⚠️ compatibility path or limited validation, ❌ unsupported.
+
+| Feature | DLO + AllGather | DLO without AllGather |
+| --- | --- | --- |
+| **DP** | ✅ Primary path; host weights are sharded across the DP group. | ✅ Each DP rank streams complete rank-local blocks. |
+| **SP** | ✅ When DP=1, DLO uses the SP group for weight sharding. | ✅ SP remains active without a DLO weight collective. |
+| **TP > 1** | ⚠️ Ordinary TP-aware loader only; no direct checkpoint mmap. | ⚠️ Ordinary TP-aware loader only; no direct checkpoint mmap. |
+| **HSDP** | ❌ Rejected to avoid double-sharding parameters. | ⚠️ Limited end-to-end coverage. |
+| **Per-tensor online FP8 linears** | ✅ Ordinary loader finalizes weights and scales before DLO sharding. | ✅ Ordinary loader retains complete rank-local tensors. |
+| **Other online quantization methods** | ❌ Rejected until runtime packing and scale layouts are validated. | ⚠️ Allowed through the ordinary loader; validation is method-specific. |
+| **Model-level or standard layerwise CPU offload** | ❌ Disabled because DLO takes priority. | ❌ Disabled because DLO takes priority. |
+| **Resident leading layers** | ❌ Rejected. | ✅ Requires eligible resident paths in the model's `OffloadPlan`. |
+
+See [Parallelism compatibility](#parallelism-compatibility) and
+[Request and loading constraints](#request-and-loading-constraints) for the
+detailed contracts and validation boundaries.
+
+## Status
+
+DLO is implemented for multi-device diffusion execution. The default
+AllGather path is the primary path for DP and SP deployments. The
+`--dlo-no-use-allgather` path streams complete blocks independently and adds no
+DLO weight collective.
+
+Host storage is selected separately from the transfer protocol. The loader can
+produce a direct-checkpoint mmap plan for a proven-compatible runtime layout;
+otherwise it uses the ordinary loader. Consequently, no-AllGather replicas on
+the same node can share immutable checkpoint pages when direct mmap is
+selected, while the ordinary-loader fallback still keeps a private runtime
+copy per process.
+
+The Phase A shared-mmap support boundary is TP1. TP greater than one is an
+ordinary-loader compatibility path: DLO can consume the resulting TP-local
+tensors, but those configurations do not use checkpoint mmap and must not be
+used to claim shared-mmap host-memory savings.
+
+The compatibility matrix below describes the current implementation. The
+unit-level guards are covered, but not every parallelism combination has a
+full model-and-hardware end-to-end test.
+
+## Design
+
+### DLO consumes the existing parallel topology
+
+DLO does not create a new DP, TP, or SP topology. It reads the configured
+`DiffusionParallelConfig` and attaches offload hooks to the DiT blocks after the
+standard distributed groups have been initialized.
+
+The DLO weight-sharding group is selected as follows:
+
+1. Use the existing DP group when `data_parallel_size > 1`.
+2. When DP is one and SP is greater than one, use the SP group.
+3. Otherwise, run rank-locally without a DLO process group.
+
+TP is deliberately not used as DLO's AllGather group. HSDP has its own
+parameter-sharding lifecycle and is not allowed to be sharded a second time by
+DLO's AllGather path.
+
+### The loader owns host-weight planning
+
+Before it decides whether ordinary weight materialization can be skipped, the
+diffusion loader builds one `HostWeightPlan`. A direct-checkpoint mmap plan is
+accepted only when preflight proves all of the following:
+
+- every required DiT parameter and persistent buffer has exactly one source;
+- runtime names, checkpoint keys, shapes, and dtypes match;
+- the runtime topology is TP1 without HSDP or online quantization; and
+- every custom loader operation is represented by a loader-owned checkpoint
+  adapter.
+
+The exact plan object is handed to DLO. The backend does not rescan checkpoint
+files, repeat the capability decision, or reconstruct names from its block
+topology. If preflight fails, the loader materializes weights normally and DLO
+consumes those runtime tensors.
+
+The plan owns only dedicated DiT component sources. If a pipeline also exposes
+ordinary sources for a text encoder or another non-DiT component, the loader
+still consumes those sources and includes their loaded names in its strict
+coverage check. Only the source prefixes covered by the plan skip ordinary
+materialization. A source that mixes DiT and non-DiT weights fails closed to
+the complete ordinary-loader path because it cannot be skipped safely as a
+unit.
+
+This boundary keeps checkpoint semantics out of DLO and avoids model-pipeline
+flags such as `_supports_mmap_loading` or parameter attributes for mmap-only
+transforms. Model-specific direct-layout knowledge, when required, lives in a
+checkpoint adapter beside the ordinary loader.
+
+### AllGather path
+
+With the default `dlo_use_allgather=True`, each rank stores approximately
+`1 / group_size` of each streamable block in pinned host memory. The next
+block's shard is copied to a device buffer and reconstructed with
+`all_gather_into_tensor` on a communication stream while the current block is
+executing.
+
+```text
+Compute:    [Block N]             [Block N+1]          [Block N+2]
+H2D:                      [shard N+1]           [shard N+2]
+AllGather:                [full N+1]             [full N+2]
+Buffers:    [current slot]       [prefetch slot]       [current slot]
+```
+
+![DLO double-buffer prefetch pipeline](../../figures/dlo/dlo_pipeline.gif)
+
+The backend uses two shared device buffers, so accelerator weight residency is
+bounded by the largest streamed blocks rather than the complete model.
+
+When direct checkpoint mmap is selected, the checkpoint mappings are only the
+source used to prepare each rank's persistent shard. They can be closed after
+shard preparation. Across the AllGather group, those private shards total
+approximately one runtime model copy.
+
+An effective DLO group size of one performs no collective, even when
+`dlo_use_allgather=True`; it follows the rank-local transfer path described
+below.
+
+When DP is greater than one, the engine can process one request per DP rank in
+the same denoising wave. Because AllGather is a collective, all participating
+requests must take the same execution path at every denoising step.
+
+### Rank-local path without DLO AllGather
+
+With `--dlo-no-use-allgather`, DLO forces its internal offload shard size to
+one and streams complete blocks using H2D copies only. The host backing may be
+either a loader-approved checkpoint mapping or ordinary runtime tensors.
+
+For direct mmap, each process retains immutable safetensors views and uses two
+bounded pinned host staging slots. Processes on the same node that map the same
+files share physical checkpoint pages through the OS page cache. This removes
+the persistent private full-model copy per pure-DP process, but each process
+still packs and transfers every complete block. Sharing is node-local; each
+node has its own page cache.
+
+When direct mmap preflight fails, the regular model loader remains responsible
+for preparing each rank's weights, including TP-local tensors or HSDP-managed
+parameters. In that fallback, each pure-DP process keeps a private full runtime
+copy.
+
+This mode means:
+
+- DP still provides independent replicas, but DLO does not shard weights
+  across DP ranks.
+- SP still performs its normal activation/attention collectives, but DLO does
+  not shard weights across SP ranks.
+- TP/HSDP/SP collectives, if configured, are not disabled by this flag; only
+  DLO's additional weight AllGather is disabled.
+- Pure DP deployments share one checkpoint-backed copy per node when direct
+  mmap is selected; the ordinary-loader fallback keeps one private runtime
+  copy per rank.
+- The scheduler does not require a synchronized DP request wave for DLO.
+
+### Component allocator-cache retention
+
+MiniMax-H3 DLO keeps one bounded PyTorch allocator cache across its staged
+encoder, DiT, and VAE component boundaries so allocations released by one
+component can be reused by the next. The DiT prefetch path and its two shared
+device buffers are unchanged.
+
+After a component is offloaded, cached-but-unallocated memory is retained only
+while it is at most 25% of device capacity and at least 5% of device capacity
+remains physically free. Crossing either bound releases the allocator cache.
+Missing allocator telemetry also releases it conservatively. Component or
+staging failure forces a release, and an out-of-memory allocation gets one
+retry after release. This is a component-local policy rather than a global
+`empty_cache` override, so the executor's unconditional shutdown cleanup is
+preserved.
+
+Retaining allocator blocks trades higher reserved and peak device memory for
+lower lifecycle latency. The benefit is device- and workload-dependent, and a
+smaller or externally loaded device may cross a bound and return to the
+ordinary per-component cache-release behavior. Current performance validation
+covers MiniMax-H3 no-AllGather DLO on CUDA. Other MiniMax-H3 DLO transfer
+topologies and platforms receive the same bounded policy but are not claimed
+performance targets; other DLO models remain unchanged unless they explicitly
+adopt it.
+
+### Final-layout Host Weight Runtime consumer
+
+> **Status:** PR2 implementation. The representation-independent identity,
+> producer, and restoration contracts landed in
+> [PR #6445](https://github.com/vllm-project/vllm-omni/pull/6445); this consumer
+> wires the final-layout BF16 path into eligible no-AllGather DLO.
+
+Final-layout Host Weight Runtime (HWR) backing is opt-in and currently applies
+only to no-AllGather DLO. Enable it with
+`--host-weight-runtime-mode preferred` (or `required`) and
+`--host-weight-runtime-root <node-local-root>`. Registration of shared mappings
+and direct asynchronous H2D remain a separate PR3 transport change.
+
+The modes express operator fallback policy, not different artifact formats:
+
+- `preferred` consumes an exact hit, but on a miss it allows canonical loading
+  followed by post-load publication. It is the normal population path.
+- `required` consumes the same exact artifact but fails startup on a miss or
+  unusable artifact. It never invokes canonical DiT fallback or post-load
+  publication and therefore cannot populate an empty store.
+
+PR2 has no separate prewarm command. Operators populate one matching producer
+cohort per node-local storage domain in `preferred` mode, then restart the same
+model revision and parallel layout with `required`. TP coordinates have distinct
+identities, while equivalent DP replicas share them.
+
+The current producer/consumer boundary is the model-declared final-layout BF16
+contract (MiniMax H3 today). It supports ordinary-loader final layouts for TP1
+and TP2 rank identities plus SP layout identities; online quantization, HSDP,
+LoRA/adapted weights, and non-default load formats remain ineligible. HWR mode
+`disabled`, DLO-disabled, and DLO AllGather configurations stop before source
+identity or store construction and retain the existing checkpoint-mmap or
+ordinary-loader path.
+
+Eligibility is decided before constructing HWR, resolving canonical sources,
+hashing identity inputs, probing a filesystem, or emitting an HWR observer
+event:
+
+```text
+DLO disabled / HWR disabled / DLO AllGather
+  -> zero final-layout HWR interaction
+  -> preserve current checkpoint-mmap or ordinary-loader behavior
+```
+
+When final-layout HWR is explicitly enabled for no-AllGather DLO, runtime
+outcomes are authoritative:
+
+- `LOCAL_HIT`: plan and commit an exact restoration, then transfer the lease
+  transactionally to DLO.
+- `CANONICAL_FALLBACK`: bypass checkpoint mmap, canonically materialize and
+  finalize the DiT, then call `publish_after_load()` for a future startup. The
+  current startup retains its canonical tensors.
+- `FAILED`: fail startup. Preferred mode does not reinterpret nonretryable
+  identity, configuration, or compatibility failures as misses.
+- Required mode cannot bootstrap an empty store through a `POST_LOAD_ONLY`
+  producer.
+
+Checkpoint mmap remains an unchanged control path whenever final-layout HWR is
+not selected.
+
+#### Pre-service transaction boundary
+
+The startup transaction does not end at restore commit. A lease-backed model is
+disposable until backend setup and initial prefetch complete:
+
+```text
+UNRESOLVED
+  -> LEASE_OWNED_BY_LOADER
+  -> RESTORE_PLANNED
+  -> COMMIT_STARTED
+  -> CARRIER_OWNED
+  -> BACKEND_OWNED
+  -> BACKEND_READY
+  -> IN_SERVICE
+```
+
+Failure before `COMMIT_STARTED` closes the loader-owned lease. Preferred mode
+may canonically load using the untouched skeleton; required mode fails.
+
+Failure from `COMMIT_STARTED` through `BACKEND_READY` must:
+
+1. synchronize or otherwise quiesce partial backend and initial-prefetch work;
+2. release hook, staging, restore-plan, and model references;
+3. close the lease through its current owner;
+4. discard the restored model; and
+5. in preferred mode, construct a fresh canonical model while bypassing HWR
+   lookup, HWR publication, and checkpoint mmap for that recovery attempt.
+
+Required mode fails instead of constructing the fresh fallback. Once the model
+enters service, failures follow normal runtime handling rather than startup
+fallback.
+
+#### Finalization phases
+
+Cold canonical loading separates byte-changing work from shared runtime-state
+finalization:
+
+```text
+Cold-only byte-changing:
+  casting, reordering, packing, quantization, generated scales,
+  model-specific weight transforms, and calibration that mutates
+  parameters or persistent buffers
+
+Shared cold/warm non-byte:
+  validation, hook installation, eval state, bookkeeping,
+  and non-persistent runtime state
+```
+
+The warm path must snapshot restored tensor bytes and backing pointers before
+shared finalization and prove both are unchanged afterward.
+
+#### Lease ownership
+
+Lease transfer is single-owner and single-take:
+
+```text
+loader owns
+  -> restore plan borrows
+  -> commit succeeds
+  -> runner carrier takes once
+  -> backend takes once before asynchronous work
+  -> backend drains pending H2D work
+  -> backend releases hooks, staging, and model references
+  -> backend closes lease
+```
+
+The runner carrier is process-local, rejects serialization, and rejects a
+duplicate `take()`. If the carrier still owns the lease, runner cleanup closes
+it. Once the backend takes ownership, only backend abort or teardown may drain
+work and close it.
+
+The implementation keeps the final-layout lease through backend setup and
+initial prefetch. The backend uses the existing two bounded rank-local staging
+slots for the immutable HWR tensors, and its transactional `enable()` cleanup
+removes partial hooks and closes the lease before reporting a setup failure.
+Preferred mode then uses the runner's fresh canonical retry; required mode
+propagates the failure.
+
+#### PR2 promotion gates
+
+- Warm hit performs zero ordinary DiT materialization and zero producer calls.
+- Shared warm finalization changes neither restored bytes nor backing pointers.
+- Preferred `FAILED` outcomes do not silently fall back.
+- Planning failure leaves the skeleton untouched.
+- Any pre-service failure after commit begins discards the restored model;
+  preferred mode uses a fresh canonical model and required mode fails.
+- Mixed components load normally and no required tensor remains on `meta`.
+- Disabled mode and AllGather emit zero final-layout HWR interaction.
+- The lease carrier rejects duplicate take and serialization.
+- Backend setup or prefetch failure drains asynchronous work before lease close.
+- Compatible checkpoint mmap remains an unchanged benchmark and control path.
+- Prewarm uses one matching producer cohort per TP/SP topology and storage
+  domain.
+- Benchmarks compare ordinary loading, cold publication, and warm staging with
+  output parity, startup latency, aggregate PSS, page-cache state, HBM, and H2D
+  payload.
+
+## Parallelism compatibility
+
+| Parallelism | DLO + AllGather | DLO without AllGather |
+| --- | --- | --- |
+| **DP** | Supported primary path. DLO shards host weights across the DP group and can run DP multi-concurrency. | Supported rank-local path. Compatible TP1 replicas can share checkpoint pages on each node; fallback runtime tensors remain private. |
+| **SP** | Supported in the implementation. With DP=1, DLO uses the SP group for host-weight sharding; SP still shards sequence/activation work. | SP remains active, but DLO keeps standard-loader rank-local weights and adds no SP weight collective. |
+| **TP > 1** | Outside the Phase A shared-mmap support scope. The loader falls back before mutation, preserves TP-local layouts, and DLO may apply DP/SP host sharding to those ordinary runtime tensors. | Outside the Phase A shared-mmap support scope. The ordinary TP-aware loader produces rank-local tensors, which DLO streams without an additional weight collective; DP replicas retain private runtime storage. |
+| **HSDP** | Rejected. HSDP has already sharded parameters, so DLO AllGather would double-shard them. | Accepted by configuration. HSDP owns parameter sharding and its own gathers; DLO only stages rank-local parameters. End-to-end coverage is limited. |
+
+### Combined dimensions
+
+- **DP + SP:** DLO uses the DP group for weight sharding when DP is greater
+  than one; SP continues to use its own sequence-parallel group. If DP is one,
+  the SP group becomes DLO's sharding group in AllGather mode.
+- **DP + TP/SP without AllGather:** standard model loading defines the
+  rank-local tensor layout. DLO adds no cross-DP, cross-TP, or cross-SP weight
+  collective. When the model declares the final-layout contract, HWR keys the
+  reusable artifact by TP rank/size and SP layout.
+- **HSDP + SP:** the general parallel configuration permits HSDP over SP, but
+  DLO must use `--dlo-no-use-allgather`. HSDP remains responsible for weight
+  materialization and synchronization.
+- **HSDP + DP or TP:** rejected independently by the diffusion parallel
+  configuration.
+
+## Request and loading constraints
+
+AllGather DP multi-concurrency requires:
+
+- explicit `num_inference_steps`;
+- the same `num_inference_steps` for all requests in a wave; and
+- identical request arguments that affect the collective execution path.
+
+The no-AllGather path does not impose these DLO-specific synchronized-wave
+requirements.
+
+Direct checkpoint mmap can back either transfer path. It is currently limited
+to proven TP1, non-HSDP, non-online-quantized layouts. Other layouts use the
+ordinary loader. Per-tensor online FP8 linears can use DLO AllGather after the
+ordinary loader finalizes their runtime weights and scales; DLO then shards and
+reconstructs those tensors with their recorded layouts. Other online methods
+must use `--dlo-no-use-allgather` or disable online quantization until their
+runtime layouts are validated.
+
+The Host Weight Runtime representation and publication contracts are merged;
+see [RFC #6414](https://github.com/vllm-project/vllm-omni/issues/6414) and
+[PR #6445](https://github.com/vllm-project/vllm-omni/pull/6445). The planned
+no-AllGather consumer above is the PR2 implementation; registration and direct
+asynchronous H2D remain a separate transport follow-up.
+
+## Validation coverage
+
+Current source-level validation includes:
+
+- HSDP + DLO + AllGather rejection;
+- HSDP + DLO without AllGather acceptance at configuration level;
+- loader preflight fallback for TP, HSDP, online quantization, unknown custom
+  loaders, missing keys, and shape/dtype mismatches;
+- ordinary-loader fallback for per-tensor online FP8 linears followed by DLO
+  sharding of finalized weights and scales;
+- exact loader-to-backend plan transfer and ordinary-loader fallback;
+- rank-local mmap source retention, bounded two-slot staging, and adapter
+  transforms without parameter-side flags;
+- resident-layer requests requiring no-AllGather;
+- DP request-wave validation for denoising-step compatibility;
+- bounded component allocator-cache retention, conservative/forced release,
+  OOM retry, and immutable encoder non-block staging;
+- sharding, double-buffer, AllGather-size, and heterogeneous-block regression
+  tests.
+
+### B300 parallel-topology smoke matrix
+
+A four-GPU B300 smoke test covered MiniMax-H3 FL2VA with the same prompt, seed,
+CUDNN attention backend, 256x256 output, two denoising steps, and
+`dlo_resident_layers=0`. The TP2 rows used DiT DP2xTP2 with the text encoder and
+VAEs at TP1. They validate the ordinary-loader fallback only, not direct mmap
+or shared-mmap host-memory savings.
+
+| Configuration | Result | Warm E2E | Peak device memory | Host PSS |
+| --- | ---: | ---: | ---: | ---: |
+| DP4xTP1 AllGather | Passed, 4 concurrent requests | 2.87 s / 4 requests | 13.84 GiB | 211.99 GiB |
+| DP4xTP1 no-AllGather | Passed, 1 request | 15.02 s | 13.23 GiB | 187.77 GiB |
+| DP2xTP2 AllGather | Passed, 2 concurrent requests | 4.16 s / 2 requests | 12.50 GiB | 211.97 GiB |
+| DP2xTP2 no-AllGather | Passed, 1 request | 3.51 s | 11.88 GiB | 314.01 GiB |
+
+Within each topology, the AllGather and no-AllGather video and audio outputs
+were byte-identical. All four runs completed without an `ERROR` or traceback
+and released their device allocations. For DP4xTP1, no-AllGather direct mmap
+reduced total PSS by 24.22 GiB (11.4%) and `Private_Dirty` from 211.33 to
+125.32 GiB (40.7%) relative to AllGather. For DP2xTP2, preflight selected the
+ordinary loader as designed; no-AllGather PSS was 314.01 GiB, about 48% above
+AllGather, because DP replicas did not share checkpoint-backed runtime
+weights. This is a functional and memory smoke test, not a production-quality
+performance or output-quality benchmark.
+
+### Host-memory measurement
+
+A two-worker MiniMax-H3 FL2VA measurement on one L20X node compared the
+ordinary-loader fallback with direct mmap. Both runs used
+DP=2, TP=1, no DLO AllGather, BF16 weights, two denoising steps, and a
+256x256 four-second request. The ordinary-loader workers were sampled after
+initialization. The mmap workers were sampled after one completed request, so
+the checkpoint working set had been faulted into the page cache; this is the
+more conservative point for mmap.
+
+The values below come from `/proc/<worker>/smaps_rollup` and include the whole
+worker, not only the DiT. The stable rank-to-rank difference comes from other
+pipeline components, so each worker should be compared with the same worker in
+the other storage mode.
+
+| Worker | Ordinary RSS | mmap RSS | Ordinary PSS | mmap PSS | PSS reduction |
+| --- | ---: | ---: | ---: | ---: | ---: |
+| DP worker 0 | 168.27 GiB | 132.76 GiB | 167.84 GiB | 101.43 GiB | 66.40 GiB |
+| DP worker 1 | 116.19 GiB | 79.97 GiB | 115.73 GiB | 48.64 GiB | 67.09 GiB |
+| **Two-worker total** | — | — | **283.56 GiB** | **150.08 GiB** | **133.48 GiB (47.1%)** |
+
+The direct-mmap workers each reported 62.45 GiB `Shared_Clean` but only
+31.20 GiB `Pss_File`, which is the proportional charge expected when the same
+resident checkpoint pages are mapped by two workers. `Private_Dirty` also fell
+from 167.53 to 70.24 GiB for worker 0 and from 115.40 to 17.44 GiB for worker
+1, a reduction of about 97–98 GiB per worker. RSS understates this benefit
+because it counts a shared physical page in every process that maps it; summed
+PSS is the appropriate node-memory comparison.
+
+The highest-value missing coverage is broader end-to-end numerical and
+lifecycle comparison against ordinary layerwise offload for DP+SP,
+HSDP+SP+no-AllGather, and TP greater than one across additional models and
+target CUDA/NCCL or CANN/HCCL hardware. That broader TP coverage does not
+change the Phase A direct-mmap TP1 support boundary.
+
+## Recommendations
+
+- Use **DP + DLO AllGather** for the supported throughput and host-memory
+  scaling path.
+- Use **SP + DLO AllGather** for long-sequence workloads when DP concurrency is
+  not the goal.
+- Use **no-AllGather** when independent replica execution is required. TP1
+  direct-mmap deployments can share checkpoint pages per node; other layouts
+  retain the ordinary loader's private host memory behavior and are outside
+  the Phase A shared-mmap support scope.
+- Prefer **HSDP alone** for production HSDP deployments until the combined
+  HSDP + DLO no-AllGather path has broader end-to-end coverage.

docs/design/feature/offloader/layerwise_offload.md

@@ -0,0 +1,57 @@
+# Layerwise Offload
+
+Layerwise offload streams transformer blocks through one accelerator. For
+commands and model support, see the
+[layerwise user guide](../../../user_guide/diffusion/offloader/layerwise_offload.md).
+
+## Hook ring
+
+`LayerWiseOffloadBackend` discovers every streamable block and registers one
+`LayerwiseOffloadHook` per block. Hooks form a ring: each hook owns the pinned
+host representation of its registered block and a reference to the next block
+it must prefetch; the final hook prefetches the first block for the next
+denoising iteration.
+
+During initialization, parameters and buffers are grouped by dtype and copied
+into contiguous pinned CPU tensors. Original tensor storage is replaced with
+an empty or meta placeholder while identity and module structure are retained.
+Metadata records names, offsets, shapes, and dtypes for rematerialization.
+
+## Forward lifecycle
+
+For each block:
+
+1. the previous hook's completion event ensures this block is materialized;
+2. the copy stream asynchronously allocates and fills the next block's device
+   tensors;
+3. parameter and buffer storage is rebound to views of those tensors;
+4. the current block computes on the normal stream; and
+5. its post-forward hook replaces device storage with placeholders after all
+   dependent work is safe.
+
+The copy stream waits on the compute stream before reusing storage, and the
+compute stream waits on the per-block prefetch event before forward. These
+dependencies, not incidental global synchronization, define correctness.
+
+## Topology contract
+
+A DiT declares ordered block-container attributes using
+`_layerwise_offload_blocks_attrs` or an `OffloadPlan`. Each resolved item must
+be an executable `nn.Module`; ordering must match forward execution. Multiple
+containers are concatenated in execution order.
+
+Non-block DiT modules, encoders, VAEs, and declared resident modules are not
+streamed by this backend.
+
+## Invariants and limitations
+
+- Each offloaded tensor has one pinned host backing store.
+- Parameter identity is preserved when storage is rebound.
+- Blocks with heterogeneous shapes and dtypes use their recorded metadata;
+  implementations must not assume identical blocks.
+- The ring is initialized before the first forward so block zero is available.
+- This backend is single-device; distributed weight reconstruction belongs to
+  [Distributed Layerwise Offload](distributed_layerwise_offload.md).
+
+Shared selection and lifecycle behavior is defined in the
+[CPU Offloading design](README.md).

docs/design/feature/omni_connectors/mooncake_store_connector.md

@@ -32,11 +32,10 @@ mooncake_master \
 
 ## Configuration
 
-Define the connector in runtime:
+Define the connector at the top level of the deploy YAML:
 
 ```yaml
-runtime:
-  connectors:
+connectors:
   connector_of_mooncake:
     name: MooncakeStoreConnector
     extra:
@@ -51,7 +50,7 @@ runtime:
 Wire stages to the connector:
 
 ```yaml
-stage_args:
+stages:
   - stage_id: 0
     output_connectors:
       to_stage_1: connector_of_mooncake

docs/design/feature/omni_connectors/mooncake_transfer_engine_connector.md

@@ -21,11 +21,10 @@ InfiniBand/RoCE NICs).
 
 ## Configuration
 
-Define the connector in runtime:
+Define the connector at the top level of the deploy YAML:
 
 ```yaml
-runtime:
-  connectors:
+connectors:
   rdma_connector:
     name: MooncakeTransferEngineConnector
     extra:
@@ -40,7 +39,7 @@ runtime:
 Wire stages to the connector:
 
 ```yaml
-stage_args:
+stages:
   - stage_id: 0
     output_connectors:
       to_stage_1: rdma_connector
@@ -415,7 +414,7 @@ A receiver connector:
 - allocates receive buffers from its own pool
 - requests metadata or transfer service from the sender
 
-The role is not inferred dynamically. It is injected by the stage configuration layer:
+The role is not inferred dynamically. It is injected from the deploy connector wiring:
 
 - incoming edge for a stage -> `role="receiver"`
 - outgoing edge for a stage -> `role="sender"`

docs/design/feature/omni_connectors/shared_memory_connector.md

@@ -13,8 +13,7 @@ segment name is returned in metadata.
 ## Configuration
 
 ```yaml
-runtime:
-  connectors:
+connectors:
   connector_of_shared_memory:
     name: SharedMemoryConnector
 ```

docs/design/feature/omni_connectors/yuanrong_connector.md

@@ -65,11 +65,10 @@ dscli stop --worker_address "${WORKER_IP}:31501"
 
 ## Configuration
 
-Define the connector in runtime:
+Define the connector at the top level of the deploy YAML:
 
 ```yaml
-runtime:
-  connectors:
+connectors:
   connector_of_yuanrong:
     name: YuanrongConnector
     extra:
@@ -81,7 +80,7 @@ runtime:
 Wire stages to the connector:
 
 ```yaml
-stage_args:
+stages:
   - stage_id: 0
     output_connectors:
       to_stage_1: connector_of_yuanrong

docs/design/feature/omni_connectors/yuanrong_transfer_engine_connector.md

@@ -94,12 +94,7 @@ For Ascend P2P, configure the connector with `protocol: "ascend"` and an NPU
 memory pool:
 
 ```yaml
-runtime:
-  enabled: true
-  defaults:
-    window_size: -1
-    max_inflight: 1
-  connectors:
+connectors:
   yuanrong_te_connector:
     name: YuanrongTransferEngineConnector
     extra:
@@ -110,10 +105,12 @@ runtime:
       device_name: "auto"
       memory_pool_size: 1073741824
       memory_pool_device: "npu"
-  edges:
+
+edges:
   - from: 0
     to: 1
     window_size: -1
+    max_inflight: 1
 ```
 
 Important fields:
@@ -141,11 +138,9 @@ For AR 4-way TP sending KV cache to DiT 4-way TP, set the stage TP topology in
 both stage `omni_kv_config` blocks:
 
 ```yaml
-stage_args:
+stages:
   - stage_id: 0
-    runtime:
     devices: "0,1,2,3"
-    engine_args:
     tensor_parallel_size: 4
     omni_kv_config:
       need_send_cache: true
@@ -156,9 +151,7 @@ stage_args:
       to_stage_1: yuanrong_te_connector
 
   - stage_id: 1
-    runtime:
     devices: "4,5,6,7"
-    engine_args:
     parallel_config:
       tensor_parallel_size: 4
     omni_kv_config:

docs/design/feature/prefix_caching.md

@@ -41,7 +41,7 @@ With this in mind, consider the set of blocks in a 2D layout, where the row repr
 
 ### Example
 !!! note "Note 3"
-    Prefix caching in vLLM-Omni currently is only supported on AutoRegressive stages with one kv-cache group. It can be enabled/disabled per-stage via the `enable_prefix_caching` parameter in the model's stage config.
+    Prefix caching in vLLM-Omni currently is only supported on AutoRegressive stages with one kv-cache group. Configure it with the pipeline-wide `enable_prefix_caching` field in the deploy config.
 
 The way in which vLLM-Omni ties into vLLM's prefix caching is best understood by example. Say that we have the following:
 

docs/design/feature/quantization.md

@@ -76,6 +76,12 @@ equivalent. `auto-round` and `auto_round` use the Omni INC/AutoRound adapter.
 ModelOpt checkpoint metadata is detected from its method, producer metadata, or
 `quant_algo`, then mapped to the corresponding vLLM ModelOpt configuration.
 
+Unless the input is a per-component dictionary, the resolved configuration is
+passed unchanged to every quantization-aware component constructed by the
+pipeline. It affects only layers supported by that component's quantization
+method; it does not rewrite arbitrary `torch.nn` modules. A per-component
+dictionary narrows this global scope through runtime layer-prefix routing.
+
 The factory uses Omni-specific overrides for methods whose diffusion behavior
 or platform dispatch is not provided by the vLLM registry. The current override
 set includes INT8, BitsAndBytes, MXFP8, MXFP4, MXFP4 dual-scale, and INC/
@@ -154,9 +160,11 @@ quantizable vLLM layers. Each method configuration is responsible for:
 Model code must keep precision-sensitive or unsupported modules unquantized.
 The shared `safe_quant_config()` helper is used by model integrations for
 normalization and modulation layers where a generic FP8 configuration would be
-unsafe. Text encoders, audio/vision encoders, VAEs, schedulers, and tokenizers
-are not implicitly quantized just because the main transformer is quantized;
-they require explicit component routing and method support.
+unsafe. Encoders exposed through the quantization factory receive a global
+configuration just like the main transformer, but only their supported vLLM
+quantizable layers are affected. Ordinary `torch.nn` encoder layers, VAEs,
+schedulers, and tokenizers are not rewritten automatically. A component map can
+select a narrower scope, but it cannot add quantization support to a layer.
 
 Weight quantization and quantized KV cache are separate paths. The former is
 resolved through `quantization_config`; runtime attention/KV-cache quantization
@@ -182,12 +190,15 @@ configured parallelism strategy:
 - `WeightsMapper` transformations must be applied before matching ignored-layer
   lists or component prefixes.
 
-Distributed layerwise offload has an additional boundary: its sharded mmap
-AllGather path is not compatible with online quantization because the path
-expects already materialized checkpoint weights. Use the rank-local
-`--dlo-no-use-allgather` path or a compatible pre-quantized checkpoint when
-combining these features. See the [DLO design document](distributed_layerwise_offload.md)
-for the full compatibility matrix.
+Distributed layerwise offload has an additional boundary: direct checkpoint
+mmap cannot prepare online-quantized tensors because it bypasses the ordinary
+loader that creates their runtime weights and scales. Per-tensor online FP8
+linears can still use DLO AllGather through the ordinary loader: each rank first
+finalizes the FP8 tensors, then DLO shards those runtime tensors for
+reconstruction during prefetch. Other online methods remain limited to
+`--dlo-no-use-allgather` until their packing and scale layouts are validated.
+See the [DLO design document](offloader/distributed_layerwise_offload.md) for
+the full compatibility matrix.
 
 ## Adding a quantization backend
 

docs/design/feature/realtime_ar_diffusion.md

@@ -0,0 +1,213 @@
+# Realtime AR-Diffusion sessions
+
+> **Status:** experimental. This document describes the internal execution
+> contract implemented under `vllm_omni.experimental.ar_diffusion`. It does not
+> define a public HTTP or WebSocket schema.
+
+## Scope
+
+Realtime AR-Diffusion serves world models that generate one autoregressive
+latent block per request while retaining model state across requests. The
+generic runtime owns identity, event ordering, session lifecycle, routing, and
+KV allocation. A model adapter owns model-specific controls and conditioning.
+
+LingBot World v2 is the first single-KV-branch integration. DreamZero uses the
+same runner capability with two logical CFG branches.
+
+## Identities
+
+Four identifiers have distinct lifetimes:
+
+| Identifier | Lifetime | Owner | Contract |
+| --- | --- | --- | --- |
+| `session_id` | One persistent world | Session manager | Selects runner KV and model-owned state. |
+| `event_id` | One control/prompt update | Transport adapter | Unique and monotonically increasing within a session, including across reset. |
+| `chunk_index` | One committed AR block position | Session | Contiguous from zero; reset starts again at zero. |
+| `request_id` | One chunk execution | Session | Correlates the tick with its returned metadata. |
+
+`AsyncOmni` keeps its normal UUID-suffixed engine routing ID. The tick
+`request_id` remains inside the immutable AR-Diffusion snapshot and the
+standard output metadata, while caller-visible `OmniRequestOutput.request_id`
+continues to use AsyncOmni's existing external-ID contract. Protocol identity
+therefore does not require a private entrypoint override.
+
+## Request lifecycle
+
+1. `ARDiffusionSession.accept_event()` validates and queues an immutable
+   `ARDiffusionSessionEvent`. Prompt and opaque control tracks may be updated in
+   the same event.
+2. `ARDiffusionSession.next_chunk()` serializes chunk execution and calls
+   `_build_tick_locked()`. The resulting `ARDiffusionTickRequest` is a snapshot
+   of the current prompt, reduced controls, ordered event IDs, and chunk index.
+3. `ARDiffusionOmniTickConsumer.execute_tick()` copies the stage sampling
+   parameters, stores the typed tick in
+   `sampling_params.extra_args["ar_diffusion_tick"]`, constructs the normal Omni
+   prompt through its `prompt_provider`, and invokes `AsyncOmni.generate()`.
+4. `AsyncOmni` allocates its normal unique engine routing ID and submits one
+   `OmniDiffusionRequest` through the orchestrator.
+5. `ARDiffusionModelRunner._request_session()` parses the typed tick.
+   `_get_or_create_session()` obtains runner-owned paged KV, and
+   `bind_ar_diffusion_state()` exposes it to the pipeline only for the duration
+   of `execute_model()`.
+6. The model pipeline validates the tick and generates exactly one latent
+   block. LingBot interprets camera controls in
+   `LingBotCameraControlReducer` and
+   `LingBotWorldCausalDMDPipeline.forward()`; the generic runtime does not parse
+   camera or action schemas.
+7. The pipeline returns one `DiffusionOutput` with the generated payload and an
+   `ARDiffusionChunkMetadata` mapping under `metadata["ar_diffusion"]`.
+   `AsyncOmni` exposes it as
+   `OmniRequestOutput.multimodal_output["metadata"]["ar_diffusion"]`.
+8. The consumer parses the standard envelope. The session commits its reducer
+   state, prompt, controls, event IDs, and next chunk index only when returned
+   metadata exactly equals the submitted tick snapshot.
+
+The metadata envelope is:
+
+```json
+{
+  "metadata": {
+    "ar_diffusion": {
+      "session_id": "world-7",
+      "request_id": "ar-world-7-3-...",
+      "chunk_index": 3,
+      "applied_event_ids": [10, 11]
+    }
+  }
+}
+```
+
+## Transaction and failure semantics
+
+Accepted events are not acknowledged as applied until the model output and
+metadata are validated. Reducer `prepare()` is speculative; reducer `commit()`
+and the session snapshot form one logical commit. A model, metadata, or reducer
+failure leaves the chunk index and queued events uncommitted and transitions the
+session to `FAILED`.
+
+After a tick failure, worker state is closed because a forward may have
+partially modified or already released KV. The same chunk is not retried in
+place. A caller must explicitly reset the session, starting again at chunk
+zero, or close it.
+
+Explicit close follows this state machine:
+
+```text
+ACTIVE or FAILED
+       |
+       v
+    CLOSING ---- worker cleanup succeeds ----> CLOSED
+       |
+       +------- worker cleanup fails --------> CLEANUP_FAILED
+                                                   |
+                                                   +-- close retry --> CLOSING
+```
+
+`CLEANUP_FAILED` is a local tombstone. The session manager retains it, rejects
+creation of the same `session_id`, and rejects reset. Only an idempotent close
+retry that succeeds removes the session. Transport disconnect uses the same
+path.
+
+## KV ownership and capacity
+
+`ARDiffusionKVCacheSpec` is the pipeline-owned immutable geometry contract. It
+declares:
+
+- TP-local self-attention geometry, AR block size, sink, and recent window;
+- logical KV branches and their worker-local slots;
+- fixed-length cross-attention caches;
+- scratch tokens required by an uncommitted forward;
+- persistent model-owned CUDA bytes per resident session; and
+- the requested resident-session capacity.
+
+The runner computes:
+
+```text
+required(capacity) =
+    scratch
+  + managed_self_attention_pages(capacity)
+  + capacity * cross_attention_bytes_per_session
+  + capacity * model_owned_state_bytes_per_session
+```
+
+`gpu_memory_fraction` is a soft expansion budget:
+
+```text
+configured_budget = available_device_bytes * gpu_memory_fraction
+effective_budget = max(configured_budget, required(1))
+```
+
+If `required(1)` exceeds actual available device memory, initialization fails.
+Otherwise the runner always admits one viable session and uses the configured
+fraction to select additional resident sessions up to the model-declared cap.
+All model-owned reservations are deducted before the paged self-attention pool
+is allocated.
+
+LingBot reports its persistent full-horizon image-condition tensor. DreamZero
+reports the measured 603 MiB per-session Wan VAE causal-convolution state.
+
+## Concurrency and routing limits
+
+The current runner executes one request at a time:
+
+- `max_num_seqs` must be one;
+- request batching and diffusion step execution are rejected;
+- one AR block is generated per engine request; and
+- an AR-Diffusion stage must have exactly one replica.
+
+The single-replica restriction is intentional. Multi-replica support requires
+session-affine routing so every tick, reset, and close reaches the worker that
+owns the session. Random or round-robin routing is not correct.
+
+Multiple resident sessions may share one runner. Capacity selection accounts
+for all persistent pools, and the runner performs LRU eviction through the same
+model lifecycle hook used by explicit close.
+
+## Model boundary
+
+The generic runtime transports controls as opaque
+`ARDiffusionControlInput(track, schema, data)` values. Model adapters validate
+and reduce those values. This keeps target/velocity, key state, pose,
+intrinsics, and other model-specific conventions out of the session protocol.
+
+For LingBot World v2:
+
+- `lingbot_world/actions.py` parses key-state and trajectory controls;
+- `lingbot_world/camera.py` loads or constructs camera trajectories and Plücker embeddings;
+- `lingbot_world/transformer.py` implements checkpoint-compatible causal attention; and
+- `lingbot_world/pipeline.py` constructs conditioning, owns small non-KV session state, and
+  produces one block plus the standard metadata envelope.
+
+## Non-goals
+
+This contract does not currently provide:
+
+- a public HTTP or WebSocket request schema (the structured-interaction
+  frontend is tracked separately in
+  [#5527](https://github.com/vllm-project/vllm-omni/pull/5527));
+- camera/action semantics shared by every world model;
+- session migration or replication across workers;
+- retry of an ambiguous partially executed chunk;
+- cross-stage KV transfer; or
+- stateful streaming VAE decode.
+
+Those features can be layered on top without adding LingBot-specific identity
+or lifecycle fields to the generic runtime.
+
+## Required regression coverage
+
+CPU contract tests cover:
+
+- separation of tick correlation IDs from internal engine routing IDs;
+- event ordering, metadata equality, and reducer fail-closed commit;
+- close/disconnect failure tombstones and cleanup retry;
+- capacity one under the default fraction for shipped LingBot geometry;
+- capacity expansion under a tuned fraction;
+- DreamZero model-owned state accounting;
+- runner session reuse, reset, close, LRU eviction, and forward cleanup; and
+- LingBot model registration, imports, camera/action reduction, and one-block
+  output metadata.
+
+GPU validation must additionally exercise real model loading, action input,
+prompt switching, rolling-window replacement, reset, close, finite output
+latents, and exact metadata/request identity.

docs/design/feature/skip_softmax.md

@@ -2,8 +2,10 @@
 
 Skip-Softmax is the sparse-attention mode of the `TRTLLM_ATTN` backend. Usage — the config keys
 and how to pick an operating point — is in
-[Diffusion Attention Backends](../../user_guide/diffusion/attention_backends.md). This page explains
-the algorithm.
+[TRTLLM Attention](../../user_guide/diffusion/attention_backends/trtllm.md#skip-softmax).
+The shared selector contract is documented in
+[Diffusion Attention Backend Selection](attention_backend_selection.md). This
+page explains the algorithm.
 
 ## Motivation
 

docs/design/index.md

@@ -18,10 +18,12 @@ implementation contract; it is not, by itself, a general support claim.
 ### Runtime and stage execution
 
 - [Disaggregated Inference](feature/disaggregated_inference.md)
+- [Host Weight Runtime](feature/host_weight_runtime.md)
 - [Async Chunk](feature/async_chunk.md)
 - [Async Diffusion Output](feature/async_diffusion_output.md)
 - [Async Omni Output Materialization](feature/omni_async_output_materialization.md)
 - [Automatic Prefix Caching in Omni Models](feature/prefix_caching.md)
+- [Realtime AR-Diffusion Sessions](feature/realtime_ar_diffusion.md)
 
 ### Communication
 
@@ -53,15 +55,22 @@ implementation contract; it is not, by itself, a general support claim.
 #### Attention optimization
 
 The [Diffusion Attention Backends](../user_guide/diffusion/attention_backends.md)
-guide lists all selectable backends and their platform defaults. Skip-Softmax
-is the backend-related optimization with a standalone feature-design contract:
+guides list selectable backends, platform defaults, installation, and tuning.
+The design contracts separate selection mechanics from backend algorithms:
 
+- [Attention Backend Selection](feature/attention_backend_selection.md)
 - [Skip-Softmax](feature/skip_softmax.md)
 
+#### CPU offloading
+
+- [Overview and Shared Contracts](feature/offloader/README.md)
+- [Model-Level Offload](feature/offloader/cpu_offload.md)
+- [Layerwise Offload](feature/offloader/layerwise_offload.md)
+- [Distributed Layerwise Offload](feature/offloader/distributed_layerwise_offload.md)
+
 - [Cache-DiT](feature/cache_dit.md)
 - [TeaCache](feature/teacache.md)
 - [Diffusion Continuous Batching](feature/diffusion_continuous_batching.md)
-- [Distributed Layerwise Offload](feature/distributed_layerwise_offload.md)
 
 ## Infrastructure and Performance
 
@@ -88,6 +97,7 @@ is the backend-related optimization with a standalone feature-design contract:
     - [Offloader](module/diffusion/offloader.md)
 - [Execution Platforms](module/execution_platforms.md)
 - [Cache Management](module/cache_management.md)
+- [Host Weight Runtime](module/host_weight_runtime.md)
 - [Quantization](module/quantization.md)
 - [Observability](module/observability.md)
 - [Profiling](module/profiling.md)

docs/design/module/diffusion/offloader.md

@@ -18,7 +18,7 @@ validation_paths:
   - tests/diffusion/offloader/**
 upstream_refs:
   - torch.nn.Module.to
-last_reviewed: 2026-07-16
+last_reviewed: 2026-08-25
 ---
 
 # Diffusion offloader
@@ -48,6 +48,11 @@ and correctness required by the selected execution mode.
 **Rule:** Retained host and device copies MUST respect configured limits and
 provide deterministic teardown.
 
+Allocator-cache retention MUST remain local to an explicit component owner,
+declare both cache and physical-free-memory bounds, release conservatively
+when telemetry is unavailable, and force release on failure or memory
+pressure. It MUST NOT replace unconditional executor shutdown cleanup.
+
 ## Safe-change guide
 
 Test enabled and disabled paths, repeated execution, asynchronous transfer,

docs/design/module/host_weight_runtime.md

@@ -0,0 +1,281 @@
+---
+title: Host Weight Runtime
+kind: module
+status: draft
+owners:
+  - "@hsliuustc0106"
+primary_code_paths:
+  - vllm_omni/host_weight_runtime/**
+related_code_paths:
+  - vllm_omni/diffusion/model_loader/**
+  - vllm_omni/diffusion/offloader/**
+depends_on:
+  - model_integration.md
+  - quantization.md
+validation_paths:
+  - tests/host_weight_runtime/**
+upstream_refs:
+  - https://github.com/vllm-project/vllm-omni/issues/6414
+last_reviewed: 2026-08-21
+---
+
+# Host Weight Runtime
+
+The Host Weight Runtime is a loader-adjacent, transport-independent foundation
+for reusing immutable, runtime-ready model weights. It separates the semantic
+question—"are these exactly the weights this loader requested?"—from the
+physical question—"where is the validated host representation stored?"
+
+The runtime can save repeated checkpoint loading, transformation, and peak host
+memory when several workers need the same final representation. It also creates
+a second, versioned path for reconstructing a valid model, so a cache miss or
+recoverable store failure normally falls back to the canonical loader.
+
+This module does not enable host-weight caching for any model by itself. A model
+integration must still provide an exact identity, a producer, and a restorer.
+For feature behavior and consumer integration requirements, see the
+[Host Weight Runtime feature design](../feature/host_weight_runtime.md).
+
+## Component boundary
+
+```mermaid
+flowchart LR
+    LA["Model loader adapter"] -->|"exact identity + policy"| RT["HostWeightRuntime"]
+    RT -->|"lookup / get_or_build"| ST["HostWeightStore"]
+    PR["WeightProducer"] -->|"runtime-ready tensors"| ST
+    ST -->|"validated mmap views"| LE["HostWeightLease"]
+    LE --> RE["WeightRestorer"]
+    RE --> MO["Restored host model"]
+    MO --> TR["DLO or another GPU transport"]
+
+    CS["Canonical model source"] --> PR
+    LA -. "fallback" .-> CS
+```
+
+The contracts have deliberately narrow ownership:
+
+- `HostWeightRuntime` applies resolution and fallback policy and emits one
+  terminal resolution report. Explicit post-load publication emits a separate
+  report and closes its store-returned lease before returning; it does not
+  revise the completed resolution.
+  The runtime does not own a model or call the canonical loader.
+- `HostWeightStore` owns artifact lookup, coordinated construction,
+  publication, validation, quarantine, and lifecycle.
+- `HostWeightLease` owns process-local tensor views, mapped-file resources, and
+  the shared artifact lock that stabilizes their lifetime.
+- `WeightProducer` creates one declared final-layout representation. It writes
+  only through a store-scoped writer and cannot publish paths directly.
+- `WeightRestorer` validates a lease-to-model plan without mutation. Its
+  one-shot `commit() -> None` is the sole model-mutating phase. Concrete
+  restorers remain with the model or quantization integration.
+- GPU transport owns page registration, private staging, H2D scheduling, and
+  lease release. DLO must not parse manifests or manage store artifacts.
+
+## Exact semantic identity
+
+Storage lookup is exact; the store never coerces one representation into
+another. `WeightArtifactIdentity` includes:
+
+- immutable model revision and source-file fingerprint;
+- component and ownership boundary;
+- runtime representation, dtype, and representation metadata;
+- final tensor layout and semantic parallel coordinates;
+- static adaptation identity; and
+- producer version, implementation fingerprint, manifest schema, and restorer
+  schema.
+
+The identity excludes transfer policy such as registered mmap versus private
+pinned staging. Those paths move the same host representation and must not
+produce duplicate artifacts.
+
+Parallel coordinates follow semantics rather than process topology:
+
+- TP size and rank are included when they change tensor contents or shape.
+- DP rank is excluded for replicated weights.
+- SP size and backend are included when they affect layout. SP rank is excluded
+  only when all SP ranks consume an identical representation.
+- PP, EP, or future parallel modes must be added by the model adapter whenever
+  they change component ownership or tensor layout.
+
+Dynamic LoRA overlays are not cached as base weights. A statically merged
+adapter is a distinct artifact identity with its own fingerprint.
+
+## Deterministic resolution
+
+The loader first resolves the canonical source and computes the exact requested
+identity. The runtime then follows this order:
+
+```text
+validated local artifact
+    -> exact remote artifact materialized locally (future)
+    -> registered local producer
+    -> canonical loader fallback, if policy permits
+```
+
+An already published local artifact wins because it avoids network transfer and
+is already in node-accessible storage. "Wins" means the highest-priority valid
+representation, not whichever backing responds first.
+
+Configuration modes are:
+
+- `disabled`: return the canonical-direct outcome without probing a root,
+  credentials, topology, or identity;
+- `preferred`: recoverable store failures may return canonical fallback; and
+- `required`: fail when no exact lease can be acquired.
+
+Failure-to-action policy is centralized in `HostWeightRuntime`. A clean miss or
+invalid cache artifact may fall back in preferred mode. A typed failure may do
+so only when it is marked retryable; unsupported capabilities, identity
+collisions, producer failures, and publication failures remain visible as
+startup failures.
+
+`coordination_timeout_seconds` bounds acquisition of lookup and build locks; it
+is not a hard wall-clock deadline for in-process validation, production, or
+atomic publication. Once this V1 implementation becomes the producer, the
+synchronous producer and publication run to completion. A hung producer blocks
+its owning process and requires external process supervision; enforceable
+producer cancellation requires a future process-isolated producer contract. A
+waiter timing out never cancels another process's valid build.
+
+Remote providers are represented by protocols but are explicitly unsupported
+in the first implementation. Pre-load `resolve()` invokes only
+`PRE_LOAD_SAFE` producers. A loader that completed canonical materialization
+may explicitly call `publish_after_load()` with a `POST_LOAD_ONLY` producer
+when `allow_post_load_publish` is enabled.
+
+`allow_local_build` controls producer use during pre-load resolution;
+`allow_post_load_publish` independently controls the explicit post-load entry
+point. This permits lookup-only warm starts that populate a missing artifact
+only after the canonical loader has established the current model.
+
+Post-load publication is synchronous in V1, but its outcome is independent of
+the already completed resolution. A store failure is returned in a separate
+publication report and cannot turn a valid canonical model into a failed
+resolution. On success, the runtime closes the validated store lease before
+returning the report. Post-load publication only warms future startups; it does
+not restore, rebind, or otherwise mutate the canonically loaded model serving
+the current startup.
+
+## Restoration transaction boundary
+
+`WeightRestorer.plan_restore()` is validation-only: it must not modify the model
+or lease. The returned plan may be committed exactly once, `commit()` returns
+`None`, and commit is the sole phase allowed to mutate the model. Registration,
+staging, and GPU transfer remain transport responsibilities rather than restore
+side effects.
+
+If planning fails, preferred-mode canonical fallback may reuse the untouched
+model. If commit begins and then fails, the model is considered partially
+hydrated and must be discarded; canonical fallback must construct a fresh model
+instance instead of reusing it.
+
+## Filesystem domain and artifact lifecycle
+
+A storage domain describes physical locality and policy; it is not part of
+semantic identity. The first implementation supports node- or NUMA-scoped
+roots on verified local disk filesystems or tmpfs. The detected kernel
+filesystem type is preserved in `domain.json` and inspection output. Known
+remote filesystems and unknown filesystem types are rejected rather than being
+silently treated as local disk, because cross-node page cache and advisory-lock
+semantics do not satisfy this backend's contract. A NUMA root alone does not
+guarantee page placement; CPU pinning, local first-touch/prefault, and residency
+evidence belong to a later topology-aware integration.
+
+```text
+<domain>/
+├── domain.json
+├── domain-policy.json
+├── locks/
+│   ├── <identity>.build.lock
+│   └── <identity>.artifact.lock
+├── artifacts/<identity>/
+│   ├── READY.json
+│   ├── manifest.json
+│   └── *.safetensors
+├── tmp/
+├── quarantine/
+└── deny/
+```
+
+Lock files live outside replaceable artifact directories. Every cooperative
+mutation follows one lock order: build lock, then exclusive artifact lock.
+Lookup acquires a shared artifact lock, validates `READY` and the manifest,
+opens payloads with non-following descriptor-based access, creates mmap-backed
+views, and transfers that shared lock to the lease.
+
+Lease acquisition rechecks the producer and restorer schemas against the exact
+identity. Malformed safetensors payloads are reported as typed invalid
+artifacts and denied rather than escaping the resolution policy as raw parser
+exceptions.
+
+Each `FileLock` records its creator PID. A forked child closes an inherited lock
+descriptor without issuing `LOCK_UN`, because `flock` state belongs to the
+shared open-file description and unlocking in the child would also release the
+parent's active lease lock.
+
+Publication follows this lifecycle:
+
+```text
+ABSENT -> invisible temporary build -> fully hashed and fsynced
+       -> atomic rename -> READY
+
+READY -> denied -> quarantined -> explicit removal
+```
+
+`READY.json` is written last in the temporary directory. Readers therefore see
+either a complete validated artifact or a miss, never an in-progress build.
+Failed process-local builds are left invisible and a later build owner removes
+their stale temporary directories. Explicit cleanup is nonblocking and reports
+an active builder or lease instead of unlinking live mappings.
+
+## Validation and capacity
+
+Every publication computes full file and tensor hashes. Local lookup supports:
+
+- `manifest_and_metadata`, which validates identity, immutable publication
+  metadata, file descriptors, sizes, tensor keys, shapes, dtypes, and strides;
+  and
+- `full_checksum`, which additionally reads and hashes every payload file.
+
+Filesystem verity is modeled but unsupported. Invalid artifacts receive an
+external deny marker before their shared lookup lock is released; a build owner
+then quarantines the immutable directory under the exclusive lock.
+Lookup checks the marker both before validation and after constructing the
+lease so a weaker overlapping lookup cannot return a lease after a stronger
+validation has denied the artifact.
+
+`inspect_domain()` and `inspect_artifact()` return typed, read-only snapshots.
+Inspection validation does not create a deny marker. Reported lock activity is
+point-in-time evidence only; kernel locks are not a persistent owner registry.
+
+Capacity policy covers all store-owned bytes, including ready, temporary, and
+quarantined data. The local writer preflights `max_artifact_bytes`,
+`max_store_bytes`, and `min_free_bytes`, and preallocates payloads where the
+filesystem supports it. `ENOSPC` remains a normal typed store failure because
+concurrent preflight is inherently racy. Automatic eviction and strict
+concurrent reservations are not implemented.
+
+For tmpfs, artifact bytes are host-memory consumption and may consume swap;
+they must not be reported operationally as ordinary disk capacity. The store
+owns files and reports capacity. The kernel controls page-cache residency, and
+transport controls registration or page locking.
+
+## First implementation scope
+
+The foundation implements a CPU local filesystem store and
+`SINGLE_PROCESS` production for one exact artifact. It intentionally adds no:
+
+- public CLI or DLO configuration;
+- BF16, FP8, or model-specific producer/restorer;
+- change to DLO AllGather or no-AllGather behavior;
+- CUDA registration, pinned staging, or H2D scheduling;
+- remote provider, group-cohort producer, artifact bundle, eviction, or
+  cross-node coordination.
+
+Consumer PRs must prove ordinary-loader parity, warm-hit work avoidance,
+parallel-layout correctness, clean teardown, shared-backing evidence, and
+representation-specific startup and memory behavior. Generic foundation tests
+remain CPU-only and make no production performance claim.
+
+The complete rationale and staged rollout are tracked in
+[RFC #6414](https://github.com/vllm-project/vllm-omni/issues/6414).

docs/design/module/model_integration.md

@@ -42,7 +42,7 @@ and execution behavior to stable runtime contracts.
 ### MODEL-INV-001: Registration is explicit
 
 **Rule:** A model integration MUST declare how its model class, loader, input
-processor, and stage configuration are selected.
+processor, and pipeline configuration are selected.
 
 ### MODEL-INV-002: Model code does not route stages
 

docs/design/module/vllm_omni_config.md

@@ -10,16 +10,15 @@ owners:
 primary_code_paths:
   - vllm_omni/config/**
   - vllm_omni/deploy/**
-  - vllm_omni/model_executor/stage_configs/**
-related_code_paths:
-  - vllm_omni/platforms/*/stage_configs/**
+  - vllm_omni/model_executor/models/*/pipeline.py
+related_code_paths: []
 depends_on: []
 validation_paths:
   - tests/config/**
 upstream_refs:
   - vllm.config
-last_reviewed: 2026-08-07
-last_verified_commit: 3d7fc3b9ba3cac88d579d4dc35b78b0b641675fc
+last_reviewed: 2026-08-11
+last_verified_commit: e356708da4cd39992405686dacfc32b18fcccc7f
 ---
 
 # vLLM-Omni configuration
@@ -39,11 +38,66 @@ The PR review proposal to capture environment-derived values in configuration
 objects at initialization is an open design question, not a current
 invariant.
 
+## Current structured-stage reuse behavior
+
+The current RFC #4021 implementation uses upstream vLLM config classes as the
+schema source for AR and generation stages:
+
+- `OmniStageLoadConfig` inherits `vllm.config.LoadConfig`.
+- `OmniStageCacheConfig` inherits `vllm.config.CacheConfig`.
+- `OmniStageSchedulerConfig` inherits `vllm.config.SchedulerConfig`.
+- `OmniStageParallelConfig` inherits `vllm.config.ParallelConfig`.
+
+This inheritance preserves the existing load, cache, scheduler, and parallel
+concern boundaries while reducing duplicate downstream declarations. It does
+not, by itself, make every inherited field an effective vLLM-Omni runtime
+option. Construction and runtime consumption are separate surfaces:
+
+| Surface | Current behavior |
+| --- | --- |
+| Structured schema | Upstream dataclass fields are inherited. Their defaults, default factories, and applicable Pydantic validation can participate when the structured object is constructed. |
+| Omni input ownership | Pipeline/deploy/CLI construction continues to accept only fields with an existing structured owner. Direct construction of an inherited sub-config can expose a wider upstream schema. |
+| Engine projection | For AR and generation stages, reusable fields are discovered by intersecting each upstream config dataclass with upstream `EngineArgs`. Only constructor-explicit values are emitted; inherited defaults remain deferred to terminal materialization. |
+| Terminal materialization | The engine-owning process constructs the final upstream `VllmConfig` and performs model-, platform-, rank-, port-, and backend-dependent initialization. |
+
+The generated projection maps include known upstream naming differences such
+as `cache_dtype` to `kv_cache_dtype`, `policy` to `scheduling_policy`, and
+`data_parallel_master_ip` to `data_parallel_address`. A field added to both an
+upstream concern config and `EngineArgs` therefore enters the AR/generation
+projection without requiring a second downstream allowlist. Explicit inherited
+fields that have no projection raise an error rather than being silently
+dropped.
+
+Ownership exclusions remain deliberate. Stage topology owns `scheduler_cls`,
+cache owns `disable_hybrid_kv_cache_manager`, and Omni runtime owns
+`distributed_executor_backend` and `worker_cls`; vLLM's private API-process
+fields are terminal internals. These exclusions prevent one input from being
+constructed or projected through two config concerns. Effective-engine-argument
+tests cover both the dynamic mapping and this exclusion boundary.
+
+`CompilationConfig` and `ProfilerConfig` are direct structured-stage fields:
+`stage.compilation_config` and `stage.profiler_config`. They accept mapping/YAML
+inputs and complete upstream value objects. Mapping inputs are validated and
+materialized as their concrete upstream types when the structured config is
+constructed, so downstream consumers see one resolved representation.
+Prebuilt upstream objects retain their type across the structured projection
+boundary.
+
+Quantization remains on its existing Omni-owned transport contract. This
+reuse step does not adopt upstream `QuantizationConfigArgs`, because the
+current engine-specific quantization split needs to be resolved separately.
+
+Diffusion keeps its existing engine-field ownership and projection surface.
+Shared Python inheritance must not be treated as evidence that an LLM-only
+parallel input is effective for a diffusion stage.
+
 ## Safe-change guide while deferred
 
 Changes should trace every affected producer to its runtime consumer and test
-the structured, legacy, CLI, and deployment paths that are actually supported.
-Do not infer a stable contract from this placeholder page.
+the structured pipeline, CLI, and deployment paths that are actually supported.
+For inherited vLLM fields, tests should separately cover constructor behavior
+and effective `EngineArgs` projection. Do not infer a stable contract from
+this placeholder page.
 
 ## Promotion gate
 

docs/design/qwen3_omni_tts_performance_optimization.md

@@ -111,7 +111,7 @@ This post walks through each optimization in the same order they are typically e
 
 For both Qwen3-Omni and Qwen3-TTS, batching is a pipeline-level optimization:
 
-- Requests are grouped per stage using `runtime.max_batch_size`
+- Requests are grouped per stage using `stages[].max_num_seqs`
 - Each stage executes batch inference with its own scheduler/worker
 - Stage outputs are routed to downstream stages with per-request mapping preserved
 
@@ -171,7 +171,7 @@ At concurrency 10, batching alone brings Qwen3-TTS RTF from 2.19 (slower than re
 
 In decode-heavy serving, repeatedly launching many small kernels from CPU can become a visible overhead. CUDA Graph reduces this overhead by capturing and replaying stable execution graphs.
 
-In stage configs, this is represented by `enforce_eager: false` for stages where graph capture is desired (Thinker/Talker), while Code2Wav keeps eager mode depending on stage behavior.
+In deploy configs, this is represented by `enforce_eager: false` for stages where graph capture is desired (Thinker/Talker), while Code2Wav keeps eager mode depending on stage behavior.
 
 ### CUDA Graph results on top of batching
 
@@ -320,7 +320,7 @@ In the async chunk pipeline, the standard `codec_chunk_frames` is 25 (each chunk
 
 **Dynamic initial chunk sizing (default behavior):**
 
-Rather than using a fixed initial chunk size, vLLM-Omni dynamically selects it based on current server load. The initial chunk size is chosen from power-of-2 steps [2, 4, 8, 16] based on load factor (`active_requests / max_batch_size`):
+Rather than using a fixed initial chunk size, vLLM-Omni dynamically selects it based on current server load. The initial chunk size is chosen from power-of-2 steps [2, 4, 8, 16] based on load factor (`active_requests / max_num_seqs`):
 
 | Server load | Initial chunk frames | Rationale |
 | --- | --- | --- |
@@ -349,8 +349,7 @@ This overrides the dynamic calculation for that request.
 **Config (server-side):**
 
 ```yaml
-runtime:
-  connectors:
+connectors:
   connector_of_shared_memory:
     name: SharedMemoryConnector
     extra:
@@ -408,7 +407,7 @@ vllm serve Qwen/Qwen3-Omni-30B-A3B-Instruct \
 
 Notes:
 
-- `runtime.max_batch_size` controls stage-level batching.
+- `stages[].max_num_seqs` controls stage-level batching.
 - Thinker/Talker commonly use `enforce_eager: false` for CUDA Graph paths.
 - Code2Wav often remains eager (`enforce_eager: true`) depending on runtime behavior.
 - Qwen3-Omni defaults `VLLM_USE_FLASHINFER_MOE_FP16=0`. The Triton has been more stable & faster
@@ -426,33 +425,26 @@ vllm serve Qwen/Qwen3-Omni-30B-A3B-Instruct \
 
 ```yaml
 async_chunk: true
-stage_args:
+stages:
   - stage_id: 0  # thinker
-    runtime:
-      max_batch_size: 64
-    engine_args:
+    max_num_seqs: 64
     enforce_eager: false
     max_num_batched_tokens: 32768
-      custom_process_next_stage_input_func: >-
-        vllm_omni.model_executor.stage_input_processors.qwen3_omni.thinker2talker_async_chunk
 
   - stage_id: 1  # talker
-    runtime:
-      max_batch_size: 64
-    engine_args:
+    max_num_seqs: 64
     enforce_eager: false
     max_num_batched_tokens: 32768
-      custom_process_next_stage_input_func: >-
-        vllm_omni.model_executor.stage_input_processors.qwen3_omni.talker2code2wav_async_chunk
 
   - stage_id: 2  # code2wav
-    runtime:
-      max_batch_size: 64
-    engine_args:
+    max_num_seqs: 64
     enforce_eager: true
     max_num_batched_tokens: 51200
 ```
 
+The async handoff processors are fixed by the registered Qwen3-Omni
+`PipelineConfig`; deploy YAML only selects `async_chunk` and runtime sizing.
+
 #### Reproduce Qwen3-Omni benchmarks
 
 ```bash
@@ -483,7 +475,7 @@ vllm-omni serve Qwen/Qwen3-TTS-12Hz-1.7B-CustomVoice \
 
 The default config (`qwen3_tts.yaml`) enables the full optimization stack:
 
-- Batching with `max_batch_size: 10` on the Talker stage
+- Batching with `max_num_seqs: 10` on the Talker stage
 - CUDA Graph on the Talker (`enforce_eager: false`)
 - Async chunk with streaming transport
 
@@ -500,25 +492,18 @@ vllm-omni serve Qwen/Qwen3-TTS-12Hz-1.7B-CustomVoice \
 
 ```yaml
 async_chunk: true
-stage_args:
+stages:
   - stage_id: 0  # Talker (AR decoder)
-    runtime:
-      max_batch_size: 10
-    engine_args:
+    max_num_seqs: 10
     enforce_eager: false
     max_num_batched_tokens: 512
-      custom_process_next_stage_input_func: >-
-        vllm_omni.model_executor.stage_input_processors.qwen3_tts.talker2code2wav_async_chunk
 
   - stage_id: 1  # Code2Wav (vocoder)
-    runtime:
-      max_batch_size: 1
-    engine_args:
+    max_num_seqs: 1
     enforce_eager: true
     max_num_batched_tokens: 8192
 
-runtime:
-  connectors:
+connectors:
   connector_of_shared_memory:
     name: SharedMemoryConnector
     extra:
@@ -527,6 +512,9 @@ runtime:
       codec_left_context_frames: 25
 ```
 
+The async Talker-to-Code2Wav processor is defined by the registered Qwen3-TTS
+pipeline rather than the deploy YAML.
+
 #### Reproduce Qwen3-TTS benchmarks
 
 ```bash

docs/features/README.md

@@ -20,13 +20,19 @@ an implementation contract has a user-facing workflow.
 | Choose serial, batched, step-wise, or streaming diffusion execution | [Execution Modes and Streaming](../user_guide/diffusion/execution_modes.md) | [Diffusion Continuous Batching](../design/feature/diffusion_continuous_batching.md), [Async Diffusion Output](../design/feature/async_diffusion_output.md) |
 | Reclaim stage memory without restarting the server | [Sleep Mode](sleep_mode.md) | Runtime lifecycle behavior is documented in the user guide |
 
+The [Prefill-Decode Disaggregation design reference](pd_disaggregation.md)
+describes the experimental Qwen3-Omni runtime topology and its current
+integration boundaries. It is not yet a supported launch recipe.
+
 Some runtime designs are deliberately not promoted as standalone User Guide
 features yet:
 
-- [Disaggregated Inference](../design/feature/disaggregated_inference.md) and
+- Generic, OmniConnector-based
+  [Disaggregated Inference](../design/feature/disaggregated_inference.md) is not
+  yet a standalone User Guide feature. Its
   [OmniConnector implementations](../design/index.md#communication) describe
-  topology and transport contracts. Practical deployment configuration remains
-  under [Pipeline and deploy configurations](../configuration/stage_configs.md).
+  transport contracts, while practical connector configuration remains under
+  [Pipeline and deploy configurations](../configuration/stage_configs.md).
 - [Async Chunk](../design/feature/async_chunk.md) and
   [Async Omni Output Materialization](../design/feature/omni_async_output_materialization.md)
   are model- and pipeline-dependent. Use the selected model's deploy
@@ -48,10 +54,10 @@ defines the shared configuration and backend extension points.
 | Goal | User guide | Related design contract |
 | --- | --- | --- |
 | Compare acceleration methods and supported combinations | [Overview](../user_guide/diffusion_features.md), [Feature Compatibility](../user_guide/feature_compatibility.md) | [Diffusion acceleration designs](../design/index.md#diffusion-acceleration) |
-| Move weights between host and device memory | [CPU Offloading](../user_guide/diffusion/cpu_offload.md) | [Distributed Layerwise Offload](../design/feature/distributed_layerwise_offload.md) |
+| Move weights between host and device memory | [CPU Offloading](../user_guide/diffusion/cpu_offload.md) | [CPU Offloading](../design/feature/offloader/README.md) |
 | Reuse denoising computation | [Cache Acceleration](../user_guide/diffusion/cache_acceleration/cache_dit.md) | [Cache-DiT](../design/feature/cache_dit.md), [TeaCache](../design/feature/teacache.md) |
 | Distribute diffusion work across devices | [Parallelism](../user_guide/diffusion/parallelism/overview.md) | [Parallelism designs](../design/index.md#parallelism) |
-| Select dense, sparse, or quantized attention paths | [Attention Backends](../user_guide/diffusion/attention_backends.md) | [Skip-Softmax](../design/feature/skip_softmax.md) |
+| Select dense, sparse, or quantized attention paths | [Attention Backends](../user_guide/diffusion/attention_backends.md) | [Attention Backend Selection](../design/feature/attention_backend_selection.md) |
 | Compile repeated diffusion regions | [Regional Compilation](../user_guide/diffusion/regional_compilation.md) | User-facing optimization guide |
 | Add generated video frames | [Frame Interpolation](../user_guide/diffusion/frame_interpolation.md) | User-facing extension guide |
 | Reduce diffusion model startup time | [Startup and Loading](../user_guide/diffusion/startup_and_loading.md) | User-facing loading guide |

docs/features/custom_pipeline.md

@@ -101,7 +101,7 @@ outputs = omni.generate(
 )
 
 # Access custom trajectory data
-output = outputs[0].request_output
+output = outputs[0]
 print(f"Trajectory timesteps shape: {output.metrics['trajectory_timesteps'].shape}")
 print(f"Trajectory latents shape: {output.latents.shape}")
 ```

docs/features/pd_disaggregation.md

@@ -0,0 +1,140 @@
+# Prefill-Decode (PD) Disaggregation (experimental)
+
+!!! warning "Experimental design reference"
+    PD disaggregation is only partially integrated on the current `main` branch.
+    The runtime contains PD detection and routing scaffolding, but there is no
+    supported deploy overlay or end-to-end validated launch recipe yet. Treat this
+    page as a description of the current design, not as production guidance.
+
+Prefill-decode disaggregation splits the Qwen3-Omni Thinker into two logical
+stages: a prefill worker that processes the prompt and produces KV cache, and a
+decode worker that imports that cache and generates tokens. The Talker and
+Code2Wav stages remain downstream; PD does not disaggregate the entire
+Qwen3-Omni pipeline.
+
+## Current design
+
+### Runtime topology
+
+```mermaid
+flowchart LR
+    client["Client request"] --> orchestrator["Omni Orchestrator"]
+    orchestrator -->|"original prompt<br/>max_tokens = 1"| prefill["Thinker prefill<br/>KV producer"]
+    prefill -->|"KV blocks"| connector[("vLLM KV connector")]
+    prefill -->|"KV routing metadata"| orchestrator
+    orchestrator -->|"original prompt<br/>remote KV parameters"| decode["Thinker decode<br/>KV consumer"]
+    connector -->|"remote KV load"| decode
+    decode -->|"generated text"| client
+```
+
+The orchestrator owns the request lifecycle, while the vLLM KV connector owns
+the bulk KV transfer. The original prompt is submitted to both Thinker stages:
+the decode worker still needs the request tokens and multimodal features to
+construct its request, while the imported KV cache prevents it from repeating
+the completed prefill computation.
+
+For the full Qwen3-Omni pipeline, splitting the Thinker adds one logical stage:
+
+| Stage | Role | Output or handoff |
+| --- | --- | --- |
+| 0 | Thinker prefill | Saves prompt KV and prompt-side multimodal output |
+| 1 | Thinker decode | Loads remote KV, generates text, and produces decode-side conditioning |
+| 2 | Talker | Converts combined Thinker conditioning into codec codes |
+| 3 | Code2Wav | Converts codec codes into audio |
+
+Text remains a final output of the decode stage, and audio remains a final
+output of Code2Wav.
+
+### Request lifecycle
+
+1. At startup, PD detection looks for one stage marked `is_prefill_only` and
+   one downstream stage marked `is_decode_only`. The decode stage must list the
+   prefill stage as an input source.
+2. The orchestrator clones the Thinker sampling parameters for prefill, forces
+   `max_tokens=1`, clears stop conditions, and sets producer-side
+   `kv_transfer_params`. This makes prefill finish by length after exporting KV.
+3. When prefill finishes, the orchestrator captures its KV routing metadata and
+   Qwen multimodal output under the original request ID.
+4. The orchestrator submits the original prompt to decode with consumer-side
+   `kv_transfer_params`. Those parameters identify the transfer, prefill engine,
+   bootstrap endpoint, and, when required by the connector, remote request.
+5. The decode engine imports the KV cache and continues generation. The normal
+   Omni stage path then carries Thinker conditioning through Talker and
+   Code2Wav.
+
+### Control and data paths
+
+PD uses two different connector layers. They solve different handoffs and are
+not interchangeable.
+
+| Path | Payload | Owner |
+| --- | --- | --- |
+| PD control path | Request ID, original prompt, `transfer_id`, bootstrap address, engine ID, and connector-specific request metadata | Omni Orchestrator |
+| PD data path | Thinker attention KV blocks | vLLM KV connector |
+| Downstream model path | Thinker hidden states and embeddings, Talker codec codes, and audio tensors | vLLM-Omni stage connectors and processors |
+
+### Detection and validation
+
+The current runtime accepts one logical prefill/decode pair per pipeline. Each
+stage may use tensor parallelism, but both sides of the pair must satisfy these
+invariants:
+
+- both stages define `kv_transfer_config`
+- prefill uses `kv_producer` or `kv_both`
+- decode uses `kv_consumer` or `kv_both`
+- both stages use the same KV connector
+- `kv_buffer_device` and `kv_buffer_size` agree when set on both stages
+- both stages use the same `tensor_parallel_size`
+
+These checks happen before request processing so an inconsistent PD topology
+fails during engine initialization.
+
+## What is implemented today
+
+| Area | Current state |
+| --- | --- |
+| Pair detection and topology validation | Implemented in the shared PD helper |
+| Prefill sampling preparation | Implemented for sync and async entrypoints |
+| Prefill-to-decode routing | Implemented in the current Orchestrator |
+| Deploy-based configuration | Not yet migrated; the supported `stages:` deploy schema cannot currently expand Qwen3-Omni into the four-stage PD topology |
+| Qwen3 Thinker-to-Talker handoff | Bridge state and embedding-merge helpers exist, but the merge is not wired into the live stage processor |
+| Connector compatibility | Incomplete; the Orchestrator currently requires `remote_request_id`, while the pinned Mooncake connector completes prefill without returning that metadata |
+| Validation coverage | PD entrypoint tests and the PD deploy overlay in online end-to-end tests are temporarily disabled during migration |
+| Performance | No non-regression or throughput/latency result is established for the current path |
+
+Because these pieces are still converging, the previous legacy `stage_args` YAML
+and `vllm serve` command have been removed. The serve CLI now accepts the
+deploy-based `stages:` schema, so retaining the legacy recipe would imply a
+launch path that current `main` rejects.
+
+## Configuration contract under development
+
+The intended deploy integration starts from the
+[default Qwen3-Omni deploy config](gh-file:vllm_omni/deploy/qwen3_omni_moe.yaml)
+and expands its frozen three-stage pipeline into the four stages shown above.
+A complete integration must:
+
+- split the Thinker definition without duplicating the Talker or Code2Wav model
+  contracts
+- remap downstream stage IDs, input sources, and stage connectors
+- inject producer and consumer `kv_transfer_config` values into the two Thinker
+  stages
+- allocate separate device sets for prefill and decode
+- preserve both final outputs: decode text and Code2Wav audio
+- provide connector-specific routing metadata without hard-coding one connector
+  contract
+
+The required number of GPUs depends on tensor-parallel sizes and whether Talker
+and Code2Wav share a device. A basic TP=1 layout uses one device for prefill, one
+for decode, and one for the two downstream audio stages.
+
+## Implementation map
+
+- PD detection, validation, and sampling-parameter preparation:
+  [`PDDisaggregationMixin`](gh-file:vllm_omni/entrypoints/pd_utils.py)
+- startup detection and connector metadata:
+  [`AsyncOmniEngine`](gh-file:vllm_omni/engine/async_omni_engine.py)
+- request-scoped prefill-to-decode routing:
+  [`Orchestrator`](gh-file:vllm_omni/engine/orchestrator.py)
+- Qwen3 downstream conditioning helpers:
+  [Qwen3-Omni stage input processor](gh-file:vllm_omni/model_executor/stage_input_processors/qwen3_omni.py)

docs/getting_started/quickstart.md

@@ -47,7 +47,7 @@ if __name__ == "__main__":
     omni = Omni(model="Tongyi-MAI/Z-Image-Turbo")
     prompt = "a cup of coffee on the table"
     outputs = omni.generate(prompt)
-    images = outputs[0].request_output.images
+    images = outputs[0].images
     images[0].save("coffee.png")
 ```
 
@@ -66,7 +66,7 @@ from vllm_omni.entrypoints.omni import Omni
 if __name__ == "__main__":
     omni = Omni(
         model="Tongyi-MAI/Z-Image-Turbo",
-        # stage_configs_path="./stage-config.yaml",  # See below
+        # deploy_config="./deploy-config.yaml",  # Optional deploy override
     )
     prompts = [
         "a cup of coffee on a table",
@@ -75,8 +75,7 @@ if __name__ == "__main__":
     ]
     omni_outputs = omni.generate(prompts)
     for i_prompt, prompt_output in enumerate(omni_outputs):
-        this_request_output = prompt_output.request_output
-        this_images = this_request_output.images
+        this_images = prompt_output.images
         for i_image, image in enumerate(this_images):
             image.save(f"p{i_prompt}-img{i_image}.jpg")
             print("saved to", f"p{i_prompt}-img{i_image}.jpg")

docs/models/supported_models.md

@@ -37,16 +37,26 @@ th {
 | `Krea2Pipeline` | Krea 2 (Raw + Turbo) | `krea/Krea-2-Raw`, `krea/Krea-2-Turbo` | ✅︎ | | | | — |
 | `WanPipeline` | Wan2.1-T2V, Wan2.2-T2V, Wan2.2-TI2V | `Wan-AI/Wan2.1-T2V-1.3B-Diffusers`, `Wan-AI/Wan2.1-T2V-14B-Diffusers`, `Wan-AI/Wan2.2-T2V-A14B-Diffusers`, `Wan-AI/Wan2.2-TI2V-5B-Diffusers` | ✅︎ | ✅︎ | ✅︎ | ✅︎ | — |
 | `WanImageToVideoPipeline` | Wan2.2-I2V | `Wan-AI/Wan2.2-I2V-A14B-Diffusers` | ✅︎ | ✅︎ | ✅︎ | ✅︎ | — |
+| `LingBotWorldCausalDMDPipeline` | LingBot-World 2.0 v1 (experimental) | `robbyant/lingbot-world-v2-14b-causal-fast-diffusers` | Experimental | | | | — |
 | `Cosmos3OmniDiffusersPipeline` | Cosmos3 T2I, T2V, I2V, V2V, T2V with sound, action policy | `nvidia/Cosmos3-Nano`, `nvidia/Cosmos3-Super` | ✅︎ | ✅︎ | ✅︎ | ✅︎ | — |
+| `SanaWmPipeline` | SANA-WM | `BBBBruce/SANA-WM_bidirectional-stage1-diffusers` | ✅︎ | | | | [Repository](https://github.com/vllm-project/vllm-omni/blob/main/recipes/NVIDIA/SANA-WM.md) |
 | `Wan22S2VPipeline` | Wan2.2-S2V | `Wan-AI/Wan2.2-S2V-14B` | ✅︎ | ✅︎ | ✅︎ | ✅︎ | — |
 | `Wan22VACEPipeline` | Wan2.1-VACE | `Wan-AI/Wan2.1-VACE-1.3B-diffusers`, `Wan-AI/Wan2.1-VACE-14B-diffusers` | ✅︎ | ✅︎ | ✅︎ | ✅︎ | — |
 | `Wan22VACEPipeline` | Wan2.2-VACE | `Pyros13/Wan2.2-VACE-Fun-A14B-Diffusers` | ✅︎ |   |   |   | — |
 | `LTX2Pipeline` | LTX-2 / LTX-2.3 one-stage T2V and I2V | `Lightricks/LTX-2`, `diffusers/LTX-2.3-Diffusers` | ✅︎ | ✅︎ | | | — |
 | `LTX2TwoStagePipeline` | LTX-2 / LTX-2.3 ordinary two-stage T2V and I2V | `Lightricks/LTX-2`, `diffusers/LTX-2.3-Diffusers` + matching Lightricks LoRA and upsampler | ✅︎ | ✅︎ | | | — |
-| `LTX2DistilledPipeline` | LTX-2 / LTX-2.3 full-distilled two-stage T2V and I2V | `rootonchair/LTX-2-19b-distilled`, `diffusers/LTX-2.3-Distilled-Diffusers` + `Lightricks/LTX-2.3` upsampler | ✅︎ | ✅︎ | | | — |
+| `LTX2DistilledOneStagePipeline` | LTX-2 / LTX-2.3 merged-distilled one-stage T2V and I2V | `rootonchair/LTX-2-19b-distilled`, `diffusers/LTX-2.3-Distilled-Diffusers` | ✅︎ | ✅︎ | | | — |
+| `LTX2DistilledTwoStagePipeline` | LTX-2 / LTX-2.3 merged-distilled two-stage T2V and I2V | `rootonchair/LTX-2-19b-distilled`, `diffusers/LTX-2.3-Distilled-Diffusers` + matching Lightricks upsampler | ✅︎ | ✅︎ | | | — |
+| `LTX2Pipeline` | LTX-2.5 Full/SFT one-stage T2V and I2V | `Lightricks/LTX-2.5-Diffusers` | ✅︎ | | | | [Repository](https://github.com/vllm-project/vllm-omni/blob/main/recipes/LTX/LTX-2.5.md) |
+| `LTX2TwoStagePipeline` | LTX-2.5 Full/SFT two-stage T2V and I2V | `Lightricks/LTX-2.5-Diffusers` + `Lightricks/LTX-2.5` LoRA | ✅︎ | | | | [Repository](https://github.com/vllm-project/vllm-omni/blob/main/recipes/LTX/LTX-2.5.md) |
+| `LTX2DistilledOneStagePipeline` | LTX-2.5 merged-distilled one-stage T2V and I2V | `Lightricks/LTX-2.5-Diffusers` | ✅︎ | | | | [Repository](https://github.com/vllm-project/vllm-omni/blob/main/recipes/LTX/LTX-2.5.md) |
+| `LTX2DistilledTwoStagePipeline` | LTX-2.5 merged-distilled two-stage T2V and I2V | `Lightricks/LTX-2.5-Diffusers` | ✅︎ | | | | [Repository](https://github.com/vllm-project/vllm-omni/blob/main/recipes/LTX/LTX-2.5.md) |
 | `LingBotVideoPipeline` | LingBot-Video dense and MoE T2I, T2V, TI2V | `robbyant/lingbot-video-dense-1.3b`, `robbyant/lingbot-video-moe-30b-a3b` | ✅︎ | | | | — |
+| `SanaVideoPipeline` | SANA-Video-2B T2V (native 480p/720p; Diffusers adapter validated at both resolutions) | `Efficient-Large-Model/SANA-Video_2B_480p_diffusers`, `Efficient-Large-Model/SANA-Video_2B_720p_diffusers` | ✅︎ | | | | [Repository](https://github.com/vllm-project/vllm-omni/blob/main/recipes/NVIDIA/SANA-Video-2B.md) |
+| `SanaImageToVideoPipeline` | SANA-Video-2B I2V (native and Diffusers adapter validated at 480p/720p) | `Efficient-Large-Model/SANA-Video_2B_480p_diffusers`, `Efficient-Large-Model/SANA-Video_2B_720p_diffusers` | ✅︎ | | | | [Repository](https://github.com/vllm-project/vllm-omni/blob/main/recipes/NVIDIA/SANA-Video-2B.md) |
 | `MiniMaxH3Pipeline` | MiniMax H3 T2VA, first/last-keyframe FL2VA, and mixed-reference Ref2VA | `MiniMaxAI/MiniMax-H3` | ✅︎ | ✅︎ | | | [Published](https://recipes.vllm.ai/MiniMaxAI/MiniMax-H3) |
 | `DreamZeroPipeline` | DreamZero-DROID | `GEAR-Dreams/DreamZero-DROID` | ✅︎ | ✅︎ | ✅︎ | ✅︎ | — |
+| `Pi0Pipeline` | π0 (Pi-Zero) VLA | `lerobot/pi0_base` | ✅︎ | | | | [Repository](https://github.com/vllm-project/vllm-omni/blob/main/recipes/lerobot/Pi0.md) |
 | `HeliosPipeline`, `HeliosPyramidPipeline` | Helios | `BestWishYsh/Helios-Base`, `BestWishYsh/Helios-Mid`, `BestWishYsh/Helios-Distilled` | ✅︎ | ✅︎ | ✅︎ | | — |
 | `MagiHumanPipeline` | MagiHuman | `SII-GAIR/daVinci-MagiHuman-Base-1080p` | ✅︎ | ✅︎ | | | — |
 | `OvisImagePipeline` | Ovis-Image | `OvisAI/Ovis-Image` | ✅︎ | ✅︎ | | ✅︎ | — |
@@ -60,17 +70,13 @@ th {
 | `OmniVoiceModel` | OmniVoice | `k2-fsa/OmniVoice` | ✅︎ | | | | — |
 | `VoxCPM2TalkerForConditionalGeneration` | VoxCPM2 | `openbmb/VoxCPM2` | ✅︎ | | | | — |
 | `DotsTTSForConditionalGeneration` | dots.tts | `rednote-hilab/dots.tts-soar` | ✅︎ | | | | — |
-| `MammothModa2ForConditionalGeneration` | MammothModa2-Preview | `bytedance-research/MammothModa2-Preview` | ✅︎ | ✅︎ | | | — |
-| `MammothModa2ForConditionalGeneration` | MammothModa2-Dev (AR-only image understanding) | `bytedance-research/MammothModa2-Dev` | ✅︎ | | | | — |
 | `Flux2KleinPipeline` | FLUX.2-klein | `black-forest-labs/FLUX.2-klein-4B`, `black-forest-labs/FLUX.2-klein-9B` | ✅︎ | ✅︎ | ✅︎ | ✅︎ | — |
 | `FluxKontextPipeline` | FLUX.1-Kontext-dev | `black-forest-labs/FLUX.1-Kontext-dev` | ✅︎ | ✅︎ | | | — |
 | `FluxPipeline` | FLUX.1-dev | `black-forest-labs/FLUX.1-dev` | ✅︎ | ✅︎ | | ✅︎ | — |
 | `FluxPipeline` | FLUX.1-schnell | `black-forest-labs/FLUX.1-schnell` | ✅︎ | ✅︎ | | ✅︎ | — |
 | `OmniGen2Pipeline` | OmniGen2 | `OmniGen2/OmniGen2` | ✅︎ | ✅︎ | | ✅︎ | — |
 | `StableAudioPipeline` | Stable-Audio-Open | `stabilityai/stable-audio-open-1.0` | ✅︎ | ✅︎ | | ✅︎ | — |
-| `SoulXSingerPipeline` | SoulX-Singer (SVS) | `Soul-AILab/SoulX-Singer` | ✅︎ | | | | — |
-| `SoulXSingerSVCPipeline` | SoulX-Singer-SVC | `Soul-AILab/SoulX-Singer` (`model-svc.pt`) | ✅︎ | | | | — |
-| `AudioXPipeline` | AudioX | `zhangj1an/AudioX` | ✅︎ | ✅︎ | | | — |
+| `MiniMaxMusic3ForConditionalGeneration` | MiniMax Music 3 (text-to-music) | `MiniMaxAI/MiniMax-Music3` | ✅︎ | | | | — |
 | `Qwen3TTSForConditionalGeneration` | Qwen3-TTS-12Hz-1.7B-CustomVoice | `Qwen/Qwen3-TTS-12Hz-1.7B-CustomVoice` | ✅︎ | | | | [Published](https://recipes.vllm.ai/Qwen/Qwen3-TTS-12Hz-1.7B-CustomVoice) |
 | `Qwen3TTSForConditionalGeneration` | Qwen3-TTS-12Hz-1.7B-VoiceDesign | `Qwen/Qwen3-TTS-12Hz-1.7B-VoiceDesign` | ✅︎ | ✅︎ | ✅︎ | ✅︎ | — |
 | `Qwen3TTSForConditionalGeneration` | Qwen3-TTS-12Hz-1.7B-Base | `Qwen/Qwen3-TTS-12Hz-1.7B-Base` | ✅︎ | ✅︎ | ✅︎ | ✅︎ | — |
@@ -98,7 +104,9 @@ th {
 |`DyninOmniForConditionalGeneration` | Dynin-Omni | `snu-aidas/Dynin-Omni` | ✅︎ | | | | — |
 | `MiniCPMO45OmniForConditionalGeneration` | MiniCPM-o 4.5 | `openbmb/MiniCPM-o-4_5` | ✅︎ | | ✅︎ | | — |
 | `ErnieImagePipeline` | ERNIE-Image | `baidu/ERNIE-Image`, `baidu/ERNIE-Image-Turbo` | ✅︎ | ✅︎ | ✅︎ | ✅︎ | — |
+| `GepardTalkerForConditionalGeneration` | Gepard-1.0 | `nineninesix/gepard-1.0` | ✅︎ | | | | — |
 |`HiDreamImagePipeline` | HiDream-I1-Full | `HiDream-ai/HiDream-I1-Full` | ✅︎ | ✅︎ | | | — |
+| `HiDreamO1ImagePipeline` | HiDream-O1-Image | `HiDream-ai/HiDream-O1-Image` | ✅︎ | | | | — |
 
 The `Recipe` column links the corresponding deployment recipe. `Published` is
 preferred when a corresponding `recipes.vllm.ai` page exists; otherwise the
@@ -111,6 +119,10 @@ support metadata until their recipe is audited.
 
 ✅︎ indicates the model is supported on that backend. Empty cells mean not listed as supported on that backend.
 
+See the [LingBot-World 2.0 recipe](https://github.com/vllm-project/vllm-omni/blob/main/recipes/Robbyant/LingBot-World-2.0.md)
+for offline and experimental realtime usage, supported controls, validation,
+and current limitations.
+
 <sup>H3</sup> MiniMax H3 AMD GPU support is validated on gfx942 (MI300X) and
 gfx950 (MI350) in BF16, using the AITER `FLASH_ATTN` diffusion attention backend.
 Other AMD SKUs (e.g. MI325X) are not listed until their

docs/serving/image_edit_api.md

@@ -123,7 +123,14 @@ When `stream=false` or omitted, the endpoint returns the standard image edit res
     }
   ],
   "output_format": null,
-  "size": null
+  "size": null,
+  "metrics": {
+    "stage_durations": {
+      "queue_wait_ms": 1.0,
+      "stage_0_gen_ms": 2.0
+    },
+    "peak_memory_mb": 1024.0
+  }
 }
 ```
 

docs/serving/speech_api.md

@@ -131,7 +131,7 @@ Content-Type: application/json
 
 | Parameter | Type | Default | Description |
 |-----------|------|---------|-------------|
-| `ref_audio` | string | null | Reference audio (HTTP URL, base64 data URL, or `file://` URI with `--allowed-local-media-path`) |
+| `ref_audio` | string | null | Reference audio (HTTP URL, base64 data URL, or `file://` URI with `--allowed-local-media-path`). Local files fold `mtime_ns` and `size` into cache keys to automatically reload on-disk edits; HTTP URLs and base64 URIs remain cached by string locator. |
 | `ref_text` | string | null | Transcript of reference audio |
 | `x_vector_only_mode` | bool | null | Use speaker embedding only (no ICL) |
 
@@ -141,7 +141,20 @@ The response shape depends on the streaming parameters:
 
 **Non-streaming (default).** With `stream=false` and no `stream_format`, returns the
 complete clip as binary audio data with an appropriate `Content-Type` header (e.g.
-`audio/wav`). The raw-bytes body has no JSON carrier, so no `usage` is reported.
+`audio/wav`). Because the raw-bytes body has no JSON carrier, successful
+non-streaming responses from non-diffusion speech servers report usage through
+response headers:
+
+| Header | Description |
+| --- | --- |
+| `x-vllm-omni-input-tokens` | Total input tokens (`text_tokens` + `audio_tokens`). |
+| `x-vllm-omni-output-tokens` | Generated codec/audio tokens. |
+| `x-vllm-omni-total-tokens` | `input_tokens` + `output_tokens`. |
+| `x-vllm-omni-input-text-tokens` | Tokens from the synthesized text (`input` plus `instructions`). |
+| `x-vllm-omni-input-audio-tokens` | Reference-audio codec frames, non-zero only for in-context voice cloning. |
+
+Diffusion-mode speech servers route through a separate response path and do not
+emit these headers.
 
 **Raw audio stream** (`stream_format="audio"`). Streams raw audio bytes (PCM or
 WAV) as they are decoded.
@@ -675,7 +688,7 @@ for result in response.json()["results"]:
 |-----------|--------|---------|-------------|
 | `tts_batch_max_items` | engine kwarg | 32 | Maximum number of items per batch request |
 
-All items are fanned out to `generate()` concurrently. The engine's stage worker automatically batches them up to the configured `max_batch_size` and queues the rest — no client-side throttling needed.
+All items are fanned out to `generate()` concurrently. The engine's stage worker automatically batches them up to the configured `max_num_seqs` and queues the rest — no client-side throttling needed.
 
 For best throughput, set both stages' `max_num_seqs` above 1 via `--stage-overrides`. On the current Qwen3-TTS CustomVoice benchmark, stage 1 performed best at `max_num_seqs: 10`:
 

docs/serving/video_stream_api.md

@@ -10,7 +10,7 @@ Each server instance runs a single model specified at startup with `vllm serve <
 
 ```bash
 vllm serve Qwen/Qwen3-Omni-30B-A3B-Instruct \
-    --deploy-config vllm_omni/deploy/qwen3_omni.yaml \
+    --deploy-config vllm_omni/deploy/qwen3_omni_moe.yaml \
     --omni \
     --port 8000 \
     --trust-remote-code

docs/serving/videos_api.md

@@ -161,11 +161,29 @@ are not implemented yet. Models may expose additional V2V controls through
 `extra_params`. For example, Cosmos3 supports
 `condition_frame_indexes_vision` and `condition_video_keep` to select which
 decoded reference frames are used as clean conditioning. Cosmos3 transfer mode
-also accepts `edge`, `blur`, `depth`, `seg`, or `wsm` control hints plus
-transfer options such as `control_path`, `control_guidance`,
-`control_guidance_interval`, `num_video_frames_per_chunk`,
-`num_conditional_frames`, `show_control_condition`, and `show_input`; see the
-Cosmos3 recipe for complete examples.
+also accepts `edge`, `blur`, `depth`, `seg`, or `wsm` control hints. Each hint
+may specify its own `control_path` and `control_weight`. Request-level transfer
+options include
+`control_guidance`, `control_guidance_interval`,
+`emphasize_control_in_prompt`, `num_video_frames_per_chunk`,
+`num_conditional_frames`, `show_control_condition`, and `show_input`. Transfer
+uses its transfer-specific system prompt and, by default, appends a
+control-adherence directive to the positive prompt. It adds duration/FPS and
+resolution metadata to both CFG branches but does not add a negative prompt
+automatically. The Cosmos3 recipe includes an optional reference negative
+prompt and shows how to pass it. Set `emphasize_control_in_prompt`,
+`use_duration_template`, or `use_resolution_template` to `false` to disable the
+corresponding addition. `negative_metadata_mode` accepts `same`, `inverse`, or
+`none` and defaults to `same` for transfer. See the Cosmos3 recipe for complete
+examples.
+
+HTTP redirects for `image_reference.image_url` follow vLLM's
+`VLLM_MEDIA_URL_ALLOW_REDIRECTS` setting. Before starting the server, set it to
+`1` (the default) to allow redirects or `0` to reject them. A redirect target
+can differ from the original host, and vLLM-Omni does not yet fully implement
+upstream vLLM's media URL allowlist protection. Deployments should therefore
+treat remote media URLs as untrusted and choose this setting as part of their
+URL access policy.
 
 ### Speech-to-Video
 

docs/user_guide/diffusion/attention_backends.md

@@ -1,63 +1,75 @@
 # Diffusion Attention Backends
 
-This document describes the diffusion attention backends available in vLLM-Omni, how to select them globally and per-role, the per-platform defaults, and how to use SageAttention.
+Use this page to select and configure a diffusion attention backend. Backend-
+specific installation, tuning, and compatibility details live in separate
+guides so that the selection contract stays easy to scan.
 
-## Overview
+Diffusion attention backend selection applies to DiT attention in image and
+video generation models. It does **not** change autoregressive LLM attention,
+which uses vLLM's attention selector.
 
-Diffusion attention backend selection is resolved in `vllm_omni.diffusion.attention.selector`. It looks up the backend from a structured `AttentionConfig` carried on `OmniDiffusionConfig` and falls back to the platform default when nothing is configured.
+For the internal selector, registry, and platform contract, see
+[Attention Backend Selection](../../design/feature/attention_backend_selection.md).
 
-This backend is used by diffusion attention layers such as the DiT attention in video and image generation models. It does **not** affect autoregressive (LLM) attention paths — those go through vLLM's own attention backend selector.
+## Choose a guide
 
-The full set of backends and their platform defaults is in the **Backend Options** and **Platform Defaults** sections below. If no attention backend is configured, vLLM-Omni asks the current platform to choose the default.
+| Need | Guide |
+| --- | --- |
+| Select a conservative or platform-native dense kernel | [Dense Backends](attention_backends/dense_backends.md) |
+| Use FlashInfer trtllm-gen FMHA, Skip-Softmax, or TRTLLM SAGE quantization | [TRTLLM Attention](attention_backends/trtllm.md) |
+| Install and use SageAttention 2.2 or SageAttention3 | [SageAttention](attention_backends/sage.md) |
+| Match training or rollout kernels loaded from Hugging Face | [Hugging Face Hub Backends](attention_backends/huggingface_hub.md) |
+| Use block-sparse video attention on Ascend NPU | [RainFusion](attention_backends/rainfusion.md) |
+| Use FastVideo VSA with FastWan2.2-TI2V-5B on CUDA | [FastVideo VSA](attention_backends/fastvideo_vsa.md) |
 
-## Backend Options
-
-| Value | Notes |
-|---|---|
-| `TRTLLM_ATTN` | FlashInfer's trtllm-gen FMHA (TensorRT-LLM's generated kernels, vendored by FlashInfer). Dense BF16, GQA native, `head_dim=128`. Datacenter Blackwell only (sm_100 / sm_103). Packed paths can provide `cu_seqlens` directly. Supports optional **Skip-Softmax** sparse attention — see below. Requires `flashinfer`. |
-| `FLASH_ATTN` | Wraps FlashAttention 4 on Blackwell when `flash-attn-4` is installed, then falls back to FlashAttention 3/2. Default on Hopper / Ada / Ampere when a compatible FlashAttention package is installed. |
-| `CUDNN_ATTN` | Pins `sdpa_kernel([CUDNN_ATTENTION])`. Default on Blackwell (sm_10x / sm_12x) with cuDNN ≥ 9.5. Wins on mask-heavy DiTs (HunyuanVideo-1.5: 2× e2e vs SDPA). |
-| `FLASHINFER_ATTN` | Uses FlashInfer's batch-prefill wrapper and supports mixed Q/K and V input dtypes through backend-specific configuration. Used as Blackwell fallback when cuDNN is unavailable. Requires `flashinfer` >= 0.6.16rc1 for mixed-dtype configurations. |
-| `TORCH_SDPA` | PyTorch `scaled_dot_product_attention` with the default backend dispatcher. Most conservative; always available. |
-| `SAGE_ATTN` | SageAttention 2.2 — INT8-quantized attention with FP16 accumulation. Lossy but typically visually indistinguishable on diffusion outputs. Requires `sageattention`. |
-| `SAGE_ATTN_3` | Requires `sageattn3` from `SageAttention/sageattention3_blackwell`. CUDA only, intended for Blackwell GPUs, with GQA/MQA requests falling back to PyTorch SDPA. |
-| `FLASH_ATTN_HUB` | FlashAttention 2 from HuggingFace `kernels` library. Useful for train/rollout alignment. |
-| `FLASH_ATTN_3_HUB` | FlashAttention 3 from HuggingFace `kernels` library. CUDA Hopper (sm_90+) only; falls back to `FLASH_ATTN_HUB` on older GPUs. |
-| `RAINFUSION_ATTN` | MindIE-SD **RainFusion** block-sparse video attention — see [below](#rainfusion_attn-backend-and-block-sparse-video-attention). Ascend NPU only; requires `mindiesd`. Delegates to `FLASH_ATTN` for anything that is not a packed video sequence. |
+## Backend options
 
+| Value | Family | Primary use | Detail |
+| --- | --- | --- | --- |
+| `TORCH_SDPA` | Dense | Conservative reference; always available | [Dense Backends](attention_backends/dense_backends.md#torch_sdpa) |
+| `FLASH_ATTN` | Dense | FlashAttention 4/3/2 depending on the installed package and GPU | [Dense Backends](attention_backends/dense_backends.md#flash_attn) |
+| `CUDNN_ATTN` | Dense | Mask-heavy DiTs on Blackwell with cuDNN 9.5 or newer | [Dense Backends](attention_backends/dense_backends.md#cudnn_attn) |
+| `FLASHINFER_ATTN` | Dense or quantized | FlashInfer batch prefill; optional mixed Q/K and V dtypes | [Dense Backends](attention_backends/dense_backends.md#flashinfer_attn) |
+| `TRTLLM_ATTN` | Dense, sparse, or quantized | Datacenter Blackwell with `head_dim=128` and compatible packed paths | [TRTLLM Attention](attention_backends/trtllm.md) |
+| `SAGE_ATTN` | Quantized | SageAttention 2.2 INT8 attention | [SageAttention](attention_backends/sage.md#sage_attn) |
+| `SAGE_ATTN_3` | Quantized | SageAttention3 on Blackwell | [SageAttention](attention_backends/sage.md#sage_attn_3) |
+| `FLASH_ATTN_HUB` | Hub kernel | FlashAttention 2 from Hugging Face `kernels` | [Hugging Face Hub Backends](attention_backends/huggingface_hub.md) |
+| `FLASH_ATTN_3_HUB` | Hub kernel | FlashAttention 3 from Hugging Face `kernels` on Hopper or newer | [Hugging Face Hub Backends](attention_backends/huggingface_hub.md) |
+| `RAINFUSION_ATTN` | Block sparse | MindIE-SD RainFusion video attention on Ascend NPU | [RainFusion](attention_backends/rainfusion.md) |
+| `FASTVIDEO_VSA` | Block sparse | FastVideo variable sparse self-attention for FastWan2.2-TI2V-5B on CUDA | [FastVideo VSA](attention_backends/fastvideo_vsa.md) |
 
 ## Configuration
 
-Diffusion attention backends can be configured three ways, in priority order:
+Backend selection is resolved in this order:
 
-1. **`--diffusion-attention-config`** — structured per-role config (highest priority).
-2. **`--diffusion-attention-backend` / `DIFFUSION_ATTENTION_BACKEND` env var** — global shorthand that sets the default backend.
-3. **Platform default** — used when nothing is configured.
+1. `--diffusion-attention-config` per-role configuration.
+2. `--diffusion-attention-backend` or `DIFFUSION_ATTENTION_BACKEND` as a
+   global default.
+3. The current platform's default.
 
-`--diffusion-attention-backend` is shorthand for `--diffusion-attention-config.default.backend`. It may be combined with `--diffusion-attention-config.per_role.*` overrides, but is mutually exclusive with `--diffusion-attention-config.default.backend`.
+`--diffusion-attention-backend` is shorthand for
+`--diffusion-attention-config.default.backend`. Do not pass it together with an
+explicit `default.backend` in the structured configuration.
 
 ### Global default
 
-Set the default backend for every diffusion attention layer:
-
 ```bash
-# CLI flag
-vllm-omni serve <model> --diffusion-attention-backend SAGE_ATTN
+vllm-omni serve <model> --diffusion-attention-backend FLASH_ATTN
 
-# Environment variable (also recognized for backwards compatibility)
-export DIFFUSION_ATTENTION_BACKEND=SAGE_ATTN
+# Backwards-compatible environment variable
+export DIFFUSION_ATTENTION_BACKEND=FLASH_ATTN
 ```
 
 ### Per-role configuration
 
-Roles are free-form strings declared by each diffusion model. The two common categories are `"self"` and `"cross"`; model-specific roles (e.g. `"ltx2.audio_to_video"`) may also be declared. A role string is matched in this order:
+Roles are declared by each diffusion model. Common categories are `self` and
+`cross`; a model may also use a more specific role such as
+`ltx2.audio_to_video`. Resolution order is:
 
-1. Exact `per_role[role]` match
-2. `per_role[role_category]` fallback (e.g. `"ltx2.audio_to_video"` → `"cross"`)
-3. `default`
-4. Platform default
-
-Use vLLM-style dotted flags or one JSON blob:
+1. Exact `per_role[role]` match.
+2. Category `per_role[role_category]` match.
+3. `default`.
+4. Platform default.
 
 ```bash
 # Dotted flags
@@ -65,430 +77,119 @@ vllm-omni serve <model> \
   --diffusion-attention-config.default.backend FLASH_ATTN \
   --diffusion-attention-config.per_role.cross.backend TORCH_SDPA
 
-# JSON
+# Equivalent JSON
 vllm-omni serve <model> \
-    --diffusion-attention-config '{"default":{"backend":"FLASH_ATTN"},"per_role":{"cross":{"backend":"TORCH_SDPA"}}}'
+  --diffusion-attention-config \
+  '{"default":{"backend":"FLASH_ATTN"},"per_role":{"cross":{"backend":"TORCH_SDPA"}}}'
 ```
 
-A backend that needs configuration exposes it as a typed field on the spec:
-`TRTLLM_ATTN`'s Skip-Softmax (see [below](#trtllm_attn-backend-and-skip-softmax)) and
-`RAINFUSION_ATTN`'s RainFusion block (see [below](#rainfusion_attn-backend-and-block-sparse-video-attention)):
-
-```bash
---diffusion-attention-config.default.backend TRTLLM_ATTN \
---diffusion-attention-config.default.skip_softmax.target_sparsity 0.5
-```
-
-### Programmatic API
-
-When constructing `OmniDiffusionConfig` directly:
+### Python API
 
 ```python
-from vllm_omni.diffusion.data import AttentionConfig, AttentionSpec, AttnQuantSpec, OmniDiffusionConfig
+from vllm_omni.diffusion.data import (
+    AttentionConfig,
+    AttentionSpec,
+    OmniDiffusionConfig,
+)
 
 config = OmniDiffusionConfig(
     diffusion_attention_config=AttentionConfig(
         default=AttentionSpec(backend="FLASH_ATTN"),
-        per_role={
-            "cross": AttentionSpec(backend="TORCH_SDPA"),
-        },
+        per_role={"cross": AttentionSpec(backend="TORCH_SDPA")},
     ),
     ...,
 )
 ```
 
-A plain dict is also accepted and normalized to `AttentionConfig`.
+Backend-specific typed blocks are documented with their consumers:
 
-#### FlashInfer QK16/V8 quantized attention
+- `quant`: [FlashInfer](attention_backends/dense_backends.md#flashinfer-quantized-attention)
+  and [TRTLLM SAGE](attention_backends/trtllm.md#sage-quantization).
+- `skip_softmax`: [TRTLLM Skip-Softmax](attention_backends/trtllm.md#skip-softmax).
+- `block_sparse`: [RainFusion](attention_backends/rainfusion.md#configuration).
+- `fastvideo_vsa_topk`: [FastVideo VSA](attention_backends/fastvideo_vsa.md#choose-top-k).
 
-`FLASHINFER_ATTN` accepts quantization options through `AttentionSpec.quant`. For QK16/V8:
-
-```python
-config = OmniDiffusionConfig(
-    diffusion_attention_config=AttentionConfig(
-        default=AttentionSpec(
-            backend="FLASHINFER_ATTN",
-            quant=AttnQuantSpec(
-                dtype_qk="bfloat16",
-                dtype_vo="fp8_e4m3",
-            ),
-        ),
-    ),
-    ...,
-)
-```
-
-`dtype_qk` configures the Q and K input dtypes. `dtype_vo` configures the V input dtype.
-
-## Platform Defaults
+## Platform defaults
 
 ### Blackwell (sm_100 / sm_103 / sm_120 / sm_121)
 
-Auto-route preference, in order:
+The CUDA auto-route preference is:
 
-1. `TRTLLM_ATTN` — on **datacenter** Blackwell (sm_100 / sm_103) when `flashinfer` is installed, the model's `head_dim` is 128, **and the model declares a compatible packed/mask-free path**
-2. `CUDNN_ATTN` — when cuDNN ≥ 9.5 is available (ships in PyTorch 2.5+ wheels)
-3. `FLASHINFER_ATTN` — when `flashinfer` is installed but cuDNN < 9.5
-4. `FLASH_ATTN` — when `flash-attn` is installed with the Blackwell CUTE kernel
-5. `TORCH_SDPA` — last resort
+1. `TRTLLM_ATTN` on datacenter Blackwell (sm_100/sm_103) when FlashInfer is
+   available, `head_dim=128`, and the model declares a compatible packed or
+   mask-free path.
+2. `CUDNN_ATTN` when cuDNN 9.5 or newer is available.
+3. `FLASHINFER_ATTN` when FlashInfer is available but cuDNN is too old.
+4. `FLASH_ATTN` when a compatible package is installed.
+5. `TORCH_SDPA`.
 
-`TRTLLM_ATTN` is skipped on workstation Blackwell (sm_120 / sm_121) and for any `head_dim != 128`, so those GPUs keep the `CUDNN_ATTN` route described below.
+`TRTLLM_ATTN` is not auto-selected on workstation Blackwell
+(sm_120/sm_121), for other head dimensions, or for paths that require an
+unsupported mask.
 
-`TRTLLM_ATTN` outranks `CUDNN_ATTN` on datacenter Blackwell for compatible packed/mask-free pipelines. Workstation Blackwell (sm_120 / sm_121) and pipelines that require attention masks retain their normal fallback.
+### Hopper, Ada, and Ampere
 
-The startup log line `Defaulting to diffusion attention backend CUDNN_ATTN (Blackwell sm_120, cuDNN 91002)` confirms the route.
+The CUDA auto-route uses `FLASH_ATTN` when available and otherwise falls back
+to `TORCH_SDPA`. `CUDNN_ATTN` and `FLASHINFER_ATTN` remain explicit options.
 
-**Why CUDNN_ATTN by default**: on mask-heavy diffusion models (HunyuanVideo-1.5, Qwen-Image), cuDNN's pinned FMHA kernel sidesteps a PyTorch SDPA dispatch quirk where the unpinned dispatcher picks `EFFICIENT_ATTENTION` (~25 ms) for masked calls instead of cuDNN (~11 ms). The pin gives 2× e2e on HV-1.5 with no regression on lighter models.
+Other platforms validate an explicit backend and choose their own default
+through the platform implementation. Check the startup log to confirm the
+resolved backend.
 
-### Hopper (sm_90) / Ada (sm_89) / Ampere (sm_80–sm_86)
+## Choosing a backend manually
 
-Auto-route preference:
+Override the platform default when you need:
 
-1. `FLASH_ATTN` — when `flash-attn` is installed
-2. `TORCH_SDPA` — fallback
+- a correctness reference (`TORCH_SDPA`);
+- a backend-specific workaround;
+- training/rollout kernel alignment (Hub backends); or
+- an explicitly validated sparse or quantized speedup.
 
-`CUDNN_ATTN` and `FLASHINFER_ATTN` are still selectable via env var on these GPUs but are not in the auto-route — FlashAttention 2 is the well-tuned path on pre-Blackwell hardware.
+The startup log prints the resolved backend and whether it came from explicit
+configuration or platform defaulting. If no resolution message appears, check
+earlier logs for diffusion-stage initialization failures.
 
-## TRTLLM_ATTN Backend and Skip-Softmax
+## Reference benchmark
 
-`TRTLLM_ATTN` runs FlashInfer's trtllm-gen FMHA. Dense it is a BF16 backend like the others; on top of
-that it can enable **Skip-Softmax**, a sparse-attention mode that trades a little fidelity for speed
-(algorithm: [design doc](../../design/feature/skip_softmax.md)).
+The following BF16 results were measured on an sm_120 RTX Pro 6000 Blackwell
+with the same prompt and seed across runs. Treat them as reference results, not
+portable guarantees.
 
-Enable it through the typed `skip_softmax` block on the attention spec:
-
-| Key | Valid values | Meaning |
-|---|---|---|
-| `target_sparsity` | finite, `[0, 1]` | Operating point on a calibrated curve. Needs a calibration for the model. |
-| `threshold` | finite, `≥ 0` | Direct threshold, no calibration needed. Mutually exclusive with `target_sparsity`. |
-| `disabled_until_timestep` | finite, `[0, 1]` | Holds early, high-noise steps dense; skip turns on once the normalized timestep `t ≤ D`. |
-
-```bash
-vllm-omni serve Wan-AI/Wan2.2-T2V-A14B-Diffusers \
-  --diffusion-attention-config '{"default": {"backend": "TRTLLM_ATTN",
-      "skip_softmax": {"target_sparsity": 0.65, "disabled_until_timestep": 0.86}}}'
-```
-
-Programmatically the same block is a typed `SkipSoftmaxSpec` (values validated at construction):
-
-```python
-from vllm_omni.diffusion.data import AttentionConfig, AttentionSpec, SkipSoftmaxSpec
-
-AttentionConfig(
-    default=AttentionSpec(
-        backend="TRTLLM_ATTN",
-        skip_softmax=SkipSoftmaxSpec(target_sparsity=0.65, disabled_until_timestep=0.86),
-    ),
-)
-```
-
-**Start at `target_sparsity=0.65, disabled_until_timestep=0.86`** and raise `target_sparsity` only
-as far as your quality bar allows — the output diverges from dense as it climbs. The gain grows with
-sequence length, since Skip-Softmax only accelerates attention. Dense BF16 is `TRTLLM_ATTN` with no
-`skip_softmax` block.
-
-Requires datacenter Blackwell with `head_dim == 128`; elsewhere, or without FlashInfer, selecting
-`TRTLLM_ATTN` raises rather than silently degrading.
-
-## TRTLLM_ATTN SAGE Quantization
-
-On top of dense/Skip-Softmax, `TRTLLM_ATTN` can run **SAGE** (SageAttention) — the trtllm-gen
-FMHA kernel with per-block quantized Q/K and per-channel fp8 V, so the two attention matmuls run
-in low precision. Enable it through the typed `quant` block on the attention spec:
-
-| Key | Valid values | Meaning |
-|---|---|---|
-| `dtype_qk` | `int8`, `fp8_e4m3` | Quantization dtype for Q and K. Absent ⇒ dense (SAGE off). |
-| `q_block_size` | `1`, `4`, `16` | Per-token block size for Q scales (default `1`). |
-| `k_block_size` | `1`, `4`, `16` | Per-token block size for K scales (default `16`). |
-
-V is always quantized per-channel to fp8_e4m3 and K-smoothing is applied inside the routine, so
-neither is a user knob. Only block sizes `1`, `4`, `16` have compiled kernels.
-
-The `quant` block is a shared `AttnQuantSpec` consumed by more than one backend; each reads the
-fields that apply and rejects the rest. `TRTLLM_ATTN` (SAGE) reads `dtype_qk` (`int8` / `fp8_e4m3`),
-`q_block_size`, and `k_block_size`; `FLASHINFER_ATTN` reads `dtype_qk` (`float16` / `bfloat16`),
-`dtype_vo`, and `flashinfer_backend` (see its section). A `dtype_qk` from the wrong set raises.
-
-```bash
-vllm-omni serve Wan-AI/Wan2.2-T2V-A14B-Diffusers \
-  --diffusion-attention-config '{"default": {"backend": "TRTLLM_ATTN",
-      "quant": {"dtype_qk": "fp8_e4m3", "q_block_size": 1, "k_block_size": 16}}}'
-```
-
-```python
-from vllm_omni.diffusion.data import AttentionConfig, AttentionSpec, AttnQuantSpec
-
-AttentionConfig(
-    default=AttentionSpec(
-        backend="TRTLLM_ATTN",
-        quant=AttnQuantSpec(dtype_qk="fp8_e4m3", q_block_size=1, k_block_size=16),
-    ),
-)
-```
-
-**Requirements.** Needs FlashInfer ≥ 0.6.16rc1. Kernel
-availability is arch-dependent: `fp8_e4m3` QK has kernels on both **SM100** (B200) and **SM103**
-(B300); `int8` QK kernels are compiled for **SM100 only**.
-
-## RAINFUSION_ATTN Backend and Block-Sparse Video Attention
-
-`RAINFUSION_ATTN` runs MindIE-SD's RainFusion (`rf_v2`) kernel on Ascend NPU. It pools each
-128-token block of keys, ranks them per query block, and computes attention against only the
-top-scoring ones. Video tokens are rearranged into `(t, h, w)` order first, so the kept blocks are
-spatiotemporal neighbours rather than arbitrary rows of the packed sequence.
-
-Only the video segment is sparsified. The prefix rows (text, visual conditions, audio) and the
-first-frame blocks are always kept dense, which is why the realized sparsity is lower than the
-nominal `sparsity` you configure. Anything the kernel cannot handle — a warmup denoise step, an
-exempt layer, a sequence with no published video geometry, a video segment under 32 blocks —
-delegates to `FLASH_ATTN`, so a model can select this backend unconditionally.
-
-Enable it through the typed `block_sparse` block on the attention spec, shared by every
-block-sparse backend:
-
-| Key | Valid values | Meaning |
-|---|---|---|
-| `sparsity` | finite, `[0, 1]` | Nominal fraction of key blocks dropped per query block. `0` disables sparsity. Defaults to `0.8`. |
-| `start_step` | `≥ 0` | Keep the first N denoise steps dense. Layout is decided early, so these steps dominate structural fidelity. |
-| `skip_layers` | index selector, e.g. `"0-3,38"` | DiT blocks that always stay dense. |
-
-```bash
-vllm-omni serve MiniMaxAI/MiniMax-H3 \
-  --diffusion-attention-config '{"default": {"backend": "RAINFUSION_ATTN",
-      "block_sparse": {"sparsity": 0.8, "start_step": 0}}}'
-```
-
-Programmatically the same block is a typed `BlockSparseSpec` (values validated at construction):
-
-```python
-from vllm_omni.diffusion.data import AttentionConfig, AttentionSpec, BlockSparseSpec
-
-AttentionConfig(
-    default=AttentionSpec(
-        backend="RAINFUSION_ATTN",
-        block_sparse=BlockSparseSpec(sparsity=0.8, start_step=0, skip_layers="0-1"),
-    ),
-)
-```
-
-**Tune in the order `start_step` → `sparsity` → `skip_layers`.** Raise `start_step` first: it is the
-cheapest way to recover structure, because the early high-noise steps decide global layout while the
-later steps only refine texture. Then walk `sparsity` up to your quality bar. Reach for
-`skip_layers` last, once an A/B against dense at the same seed points at specific blocks.
-
-Requires Ascend NPU with `mindiesd`; selecting it on any other platform raises. It is also
-incompatible with ring sequence parallelism, since `rf_v2` needs the whole key sequence to rank
-blocks — use Ulysses SP (`ring_degree=1`).
-
-### Which geometries run sparse
-
-MindIE-SD rf_v2 handles arbitrary video grids by spatially rearranging video first and, when
-necessary, promoting a real-video suffix to its always-kept prefix segment. It never pads: rf_v2
-does not consume an attention mask for padded keys. The remaining video rows stay sparse, while the
-promoted rows remain visible to every query and key.
-
-For example, MiniMax-H3 at 1344x768 has grid `(62, 24, 42)`: its 62496 video rows leave a 32-row
-128-block residue, and its two non-8-wide latent columns form a 2928-row rearranged suffix.
-MindIE-SD promotes 2976 rows, leaving 59520 sparse rows (`465 × 128`). This keeps the block mask
-and kernel tiling aligned, at the cost of retaining more blocks.
-
-8x8-aligned spatial grids remain preferable for performance. A latent `h` or `w` not divisible by 8
-creates an always-kept residual suffix and lowers realized sparsity. "Multiple of 8" on the latent
-grid means **width and height that are multiples of 256**. This path requires a MindIE-SD release
-that implements protected video tails for rf_v2.
-
-## End-to-End Benchmark (BF16, sm_120 RTX Pro 6000 Blackwell)
-
-Same prompt and seed across runs. `Total generation time` from `text_to_video.py` / `text_to_image.py`.
-
-| Model | Shape | TORCH_SDPA | CUDNN_ATTN | FLASHINFER_ATTN |
-|---|---|---|---|---|
+| Model | Shape | `TORCH_SDPA` | `CUDNN_ATTN` | `FLASHINFER_ATTN` |
+| --- | --- | ---: | ---: | ---: |
 | HunyuanVideo-1.5 (T2V) | 480p / 33f / 50 steps | 147.05 s | **73.02 s** | 127.84 s |
 | Wan 2.2 14B (T2V) | 480p / 33f / 40 steps | 117.75 s | 117.17 s | **115.07 s** |
 | Qwen-Image (T2I) | 1024² / 50 steps | 17.41 s | **15.14 s** | 16.02 s |
 | FLUX.2-dev (T2I) | 1024² / 50 steps, TP=2 | 53.62 s | **53.30 s** | 54.94 s |
 
-Pattern: mask-heavy DiTs (HV-1.5, Qwen-Image) favor `CUDNN_ATTN`; lighter-mask DiTs and TP-saturated configs (Wan 2.2, FLUX.2 TP=2) tie within noise.
+Mask-heavy DiTs favored `CUDNN_ATTN`; lighter-mask or TP-saturated workloads
+were close enough that users should benchmark their exact model and shape.
 
-## Known Limitations
+## Compatibility anchors
 
-### LTX-2.0: `CUDNN_ATTN` crashes under torch.compile
+The following headings preserve links to sections that moved into dedicated
+guides.
 
-LTX-2's audio attention has a symbolic head_dim under torch.compile tracing. cuDNN's SDPA backend selector rejects symbolic dims and Dynamo aborts compilation. Tracked in [#3121](https://github.com/vllm-project/vllm-omni/issues/3121).
+## TRTLLM_ATTN Backend and Skip-Softmax
 
-**Workaround**: explicitly select `FLASHINFER_ATTN` or `TORCH_SDPA` for LTX-2.0:
+See [TRTLLM Attention](attention_backends/trtllm.md#skip-softmax).
 
-```bash
-DIFFUSION_ATTENTION_BACKEND=FLASHINFER_ATTN python examples/offline_inference/text_to_video/text_to_video.py \
-    --model Lightricks/LTX-2 ...
-```
+## TRTLLM_ATTN SAGE Quantization
 
-### FlashAttention-4 on Blackwell
+See [TRTLLM Attention](attention_backends/trtllm.md#sage-quantization).
 
-Install the optional CUDA 13 extra to use the CuTe-based FA4 path:
+## RAINFUSION_ATTN Backend and Block-Sparse Video Attention
 
-```bash
-pip install 'vllm-omni[fa4]'
-```
-
-`FLASH_ATTN` prefers `flash_attn.cute` on Blackwell and falls back to FA3/FA2
-when it is unavailable. Version `4.0.0b18` is required; earlier beta
-wheels had known JIT failures on Blackwell.
-
-## Choosing a Backend Manually
-
-### When to override the default
-
-- **Quality validation**: compare a new backend against `TORCH_SDPA` as the reference, since SDPA's default dispatcher is the most extensively tested.
-- **Lossy speedup hunting**: try `SAGE_ATTN` (INT8 quantized) on diffusion outputs — typically indistinguishable visually but always validate.
-- **Workaround for known issues**: see Known Limitations above.
-
-### Verifying which backend is in use
-
-The startup log prints one of:
-
-```
-Using diffusion attention backend 'CUDNN_ATTN'           # explicit override
-Defaulting to diffusion attention backend CUDNN_ATTN ... # auto-route
-Defaulting to diffusion attention backend SDPA           # nothing else available
-```
-
-If you don't see one of these, the model didn't reach diffusion stage init — check earlier logs for failures.
+See [RainFusion](attention_backends/rainfusion.md).
 
 ## SageAttention Installation
 
-vLLM-Omni expects SageAttention to be installed into the same Python environment as vLLM-Omni.
-
-Build from source:
-
-```bash
-git clone https://github.com/thu-ml/SageAttention.git
-cd SageAttention
-
-export EXT_PARALLEL=4 NVCC_APPEND_FLAGS="--threads 8" MAX_JOBS=32
-pip install . --no-build-isolation
-```
-
-Quick check:
-
-```bash
-python -c "import sageattention; print(sageattention.__file__)"
-```
+See [SageAttention](attention_backends/sage.md#installation).
 
 ## SageAttention3 Installation
 
-vLLM-Omni expects SageAttention3 to be installed into the same Python environment as vLLM-Omni.
-
-Build from source:
-
-```bash
-git clone https://github.com/thu-ml/SageAttention.git
-cd SageAttention/sageattention3_blackwell
-python setup.py install
-```
-
-Quick check:
-
-```bash
-python -c "import sageattn3; print(sageattn3.__file__)"
-```
-
-Notes:
-
-- `SAGE_ATTN_3` is only selected on CUDA when `sageattn3` is importable and the GPU is Blackwell-class.
-- SageAttention3's Blackwell kernel assumes `Hq == Hkv`. In vLLM-Omni, GQA/MQA diffusion requests fall back to PyTorch SDPA for correctness.
+See [SageAttention](attention_backends/sage.md#sageattention3-installation).
 
 ## HuggingFace Kernels Hub Backends
 
-To achieve perfect numerical consistency between **training** (typically using HuggingFace Diffusers with Hub kernels) and **serving/rollout** (using vLLM-Omni), you can use the Hub-based attention backends. This eliminates numerical drift / sampling divergence caused by executing different local kernel versions during rollout.
-
-The following backend options are supported:
-- `FLASH_ATTN_HUB` (HuggingFace `kernels-community/flash-attn2`)
-- `FLASH_ATTN_3_HUB` (HuggingFace `kernels-community/flash-attn3`, Hopper sm_90+ only)
-
-### Installation
-
-To use these backends, you must install the `kernels` library:
-
-```bash
-pip install kernels==0.14.1
-```
-
-If the `kernels` library is not available in the environment, vLLM-Omni will log a warning and fall back gracefully to the corresponding local backend implementations (`FLASH_ATTN`). On CUDA GPUs below Hopper (compute capability < 9.0), `FLASH_ATTN_3_HUB` falls back to `FLASH_ATTN_HUB`.
-
-### Usage
-
-Select a Hub backend using the global CLI flag or environment variables:
-
-```bash
-# Environment variable
-export DIFFUSION_ATTENTION_BACKEND=FLASH_ATTN_HUB
-
-# CLI flag
-vllm-omni serve <model> --diffusion-attention-backend FLASH_ATTN_HUB
-```
-
-## Usage Examples
-
-### Default (auto-route)
-
-```bash
-python examples/offline_inference/text_to_video/text_to_video.py \
-    --model hunyuanvideo-community/HunyuanVideo-1.5-Diffusers-480p_t2v \
-    --prompt "A dog running across a field of golden wheat." \
-    --height 480 --width 832 --num-frames 33 \
-    --num-inference-steps 50 --seed 42 --guidance-scale 6.0 \
-    --output hv15.mp4
-```
-
-On Blackwell this picks `CUDNN_ATTN` automatically. Check the log for the `Defaulting to ...` line.
-
-### Explicit backend selection
-
-```bash
-DIFFUSION_ATTENTION_BACKEND=FLASHINFER_ATTN python examples/offline_inference/text_to_video/text_to_video.py \
-    --model Lightricks/LTX-2 \
-    --prompt "A dog running across a field of golden wheat." \
-    --height 480 --width 832 --num-frames 33 \
-    --num-inference-steps 40 --seed 42 --guidance-scale 4.0 \
-    --output ltx2.mp4
-```
-
-### SageAttention (lossy)
-
-```bash
-DIFFUSION_ATTENTION_BACKEND=SAGE_ATTN python examples/offline_inference/text_to_video/text_to_video.py \
-    --model hunyuanvideo-community/HunyuanVideo-1.5-Diffusers-480p_t2v \
-    --prompt "A dog running across a field of golden wheat." \
-    --height 480 --width 832 --num-frames 33 \
-    --num-inference-steps 30 --seed 42 --guidance-scale 6.0 \
-    --tensor-parallel-size 2 \
-    --output hv15_sage.mp4
-```
-
-Example: Wan2.2 TI2V 5B
-
-```bash
-DIFFUSION_ATTENTION_BACKEND=SAGE_ATTN python examples/offline_inference/text_to_video/text_to_video.py \
-    --model Wan-AI/Wan2.2-TI2V-5B-Diffusers \
-    --prompt "A dog running across a field of golden wheat." \
-    --height 704 --width 1280 --num-frames 49 \
-    --num-inference-steps 30 --seed 42 --guidance-scale 5.0 \
-    --tensor-parallel-size 2 \
-    --output outputs/wan22_sage.mp4
-```
-
-### Enable SageAttention3
-
-Example:
-
-```bash
-DIFFUSION_ATTENTION_BACKEND=SAGE_ATTN_3 python examples/offline_inference/text_to_video/text_to_video.py \
-    --model hunyuanvideo-community/HunyuanVideo-1.5-Diffusers-480p_t2v \
-    --prompt "A dog running across a field of golden wheat." \
-    --height 480 --width 832 --num-frames 33 \
-    --num-inference-steps 30 --seed 42 --guidance-scale 6.0 \
-    --tensor-parallel-size 2 \
-    --output outputs/hv15_sage3.mp4
-```
+See [Hugging Face Hub Backends](attention_backends/huggingface_hub.md).

docs/user_guide/diffusion/attention_backends/dense_backends.md

@@ -0,0 +1,110 @@
+# Dense Attention Backends
+
+Use these backends when you want dense attention without backend-specific
+sparsity. Start with `TORCH_SDPA` for correctness comparisons, then benchmark
+the fastest compatible kernel for your model, shape, and hardware.
+
+For selection precedence and per-role configuration, see the
+[attention backend overview](../attention_backends.md).
+
+## `TORCH_SDPA`
+
+`TORCH_SDPA` calls PyTorch `scaled_dot_product_attention` and lets PyTorch's
+dispatcher choose the implementation. It is always available and is the most
+conservative reference when validating another backend.
+
+```bash
+vllm-omni serve <model> --diffusion-attention-backend TORCH_SDPA
+```
+
+## `FLASH_ATTN`
+
+`FLASH_ATTN` uses the installed FlashAttention implementation. On Blackwell it
+prefers the CuTe-based FlashAttention 4 path and falls back to
+FlashAttention 3 or 2 when that path is unavailable. On Hopper, Ada, and
+Ampere it is the preferred automatic route when a compatible package is
+installed.
+
+```bash
+vllm-omni serve <model> --diffusion-attention-backend FLASH_ATTN
+```
+
+### FlashAttention 4 on Blackwell
+
+Install the optional CUDA 13 extra:
+
+```bash
+pip install 'vllm-omni[fa4]'
+```
+
+Version `4.0.0b18` is required; earlier beta wheels had known JIT failures on
+Blackwell. If the CuTe path is unavailable, the backend falls back to the
+compatible FlashAttention 3 or 2 path.
+
+## `CUDNN_ATTN`
+
+`CUDNN_ATTN` pins PyTorch SDPA to `CUDNN_ATTENTION`. It is particularly useful
+for mask-heavy DiTs and is automatically preferred on Blackwell when cuDNN
+9.5 or newer is available and the higher-priority TRTLLM route is not
+compatible.
+
+```bash
+vllm-omni serve <model> --diffusion-attention-backend CUDNN_ATTN
+```
+
+### LTX-2.0 limitation
+
+LTX-2 audio attention has a symbolic head dimension during `torch.compile`
+tracing. The cuDNN SDPA selector rejects that symbolic dimension and Dynamo
+aborts compilation. This is tracked in
+[issue #3121](https://github.com/vllm-project/vllm-omni/issues/3121).
+
+Use `FLASHINFER_ATTN` or `TORCH_SDPA` as a workaround:
+
+```bash
+DIFFUSION_ATTENTION_BACKEND=FLASHINFER_ATTN \
+  python examples/offline_inference/text_to_video/text_to_video.py \
+  --model Lightricks/LTX-2 ...
+```
+
+## `FLASHINFER_ATTN`
+
+`FLASHINFER_ATTN` uses FlashInfer's batch-prefill wrapper. It is an explicit
+option on CUDA platforms and an automatic Blackwell fallback when FlashInfer
+is installed but cuDNN is too old for `CUDNN_ATTN`.
+
+```bash
+vllm-omni serve <model> --diffusion-attention-backend FLASHINFER_ATTN
+```
+
+### FlashInfer quantized attention
+
+The backend accepts an `AttentionSpec.quant` block. For QK16/V8, keep Q and K
+in FP16 or BF16 and use FP8 E4M3 for V:
+
+```python
+from vllm_omni.diffusion.data import (
+    AttentionConfig,
+    AttentionSpec,
+    AttnQuantSpec,
+    OmniDiffusionConfig,
+)
+
+config = OmniDiffusionConfig(
+    diffusion_attention_config=AttentionConfig(
+        default=AttentionSpec(
+            backend="FLASHINFER_ATTN",
+            quant=AttnQuantSpec(
+                dtype_qk="bfloat16",
+                dtype_vo="fp8_e4m3",
+            ),
+        ),
+    ),
+    ...,
+)
+```
+
+`dtype_qk` controls Q and K; `dtype_vo` controls V. Mixed-dtype
+configurations require FlashInfer 0.6.16rc1 or newer. The shared quantization
+schema is also consumed by TRTLLM, but each backend validates its own allowed
+fields and values; see [TRTLLM SAGE quantization](trtllm.md#sage-quantization).

docs/user_guide/diffusion/attention_backends/fastvideo_vsa.md

@@ -0,0 +1,125 @@
+# FastVideo VSA
+
+FastVideo Variable Sparse Attention (VSA) accelerates the `FastVideo/FastWan2.2-TI2V-5B-Diffusers` model's
+self-attention by partitioning the post-patch latent grid into spatiotemporal
+blocks. For every query block, VSA scores the key/value blocks and computes
+attention only against the selected top-k blocks.
+
+The supported checkpoint provides both text-to-video and image-to-video modes through `Wan22Pipeline`; the separate Wan I2V-14B, S2V, and VACE pipelines are outside this backend's supported scope.
+
+VSA is a CUDA-only, explicitly selected backend. It requires the
+`fastvideo-kernel` package and currently supports non-causal self-attention
+with equal query and key/value sequence lengths. Unsupported shapes, masks,
+dtypes, sequence-parallel execution, or kernel failures fall back to
+`TORCH_SDPA` and emit a warning with the reason.
+
+## Enable the backend
+
+For online serving, select the backend with the existing attention backend
+flag. Use `--fastvideo-vsa-topk` to set the number of key/value blocks retained
+for every query block:
+
+```bash
+vllm-omni serve <model> \
+  --diffusion-attention-backend FASTVIDEO_VSA \
+  --fastvideo-vsa-topk 64
+```
+
+The backwards-compatible environment variable selects the backend with the
+default `topk=64`:
+
+```bash
+export DIFFUSION_ATTENTION_BACKEND=FASTVIDEO_VSA
+vllm-omni serve <model>
+```
+
+To tune top-k, pass the CLI backend and top-k flags together as shown above,
+or use the equivalent structured configuration:
+
+```bash
+vllm-omni serve <model> \
+  --diffusion-attention-config \
+  '{"default":{"backend":"FASTVIDEO_VSA","fastvideo_vsa_topk":64}}'
+```
+
+Do not combine `--diffusion-attention-backend` with an explicit
+`diffusion_attention_config.default.backend`. The top-k value must be positive
+and is valid only when the default backend is `FASTVIDEO_VSA`.
+
+For a deploy YAML stage, use either the shorthand fields:
+
+```yaml
+stages:
+  - stage_id: 0
+    diffusion_attention_backend: FASTVIDEO_VSA
+    fastvideo_vsa_topk: 64
+```
+
+or the structured configuration:
+
+```yaml
+stages:
+  - stage_id: 0
+    diffusion_attention_config:
+      default:
+        backend: FASTVIDEO_VSA
+        fastvideo_vsa_topk: 64
+```
+
+## Choose top-k
+
+At runtime the backend logs the sequence shape and derived block count:
+
+```text
+FASTVIDEO_VSA routing: seq_len=27280, dit_seq_shape=(31, 22, 40),
+block_size=(4, 8, 8), num_blocks=120, topk=64, keep_ratio=53.3%,
+checkpoint_mode=native, route=VSA
+```
+
+Use `num_blocks` as the upper bound when tuning top-k. Top-k is the number of
+key/value blocks retained for each query block, not the number of tokens or
+the total number of blocks processed by the layer.
+
+- A smaller top-k increases sparsity and may reduce attention computation, but
+  it can remove relevant blocks and reduce visual quality or temporal
+  consistency. Routing and padding overhead can also make a smaller value
+  slower for some shapes.
+- A larger top-k retains more context and generally approaches dense-attention
+  quality, but reduces the potential speedup and increases memory traffic.
+- `topk > num_blocks` is invalid for the runtime shape and falls back to SDPA.
+- For a native checkpoint, `topk == num_blocks` routes to SDPA because scoring
+  every block provides no sparsity benefit.
+- For a FastVideo DMD checkpoint, `topk == num_blocks` stays on the VSA
+  all-block path so the checkpoint keeps its trained compensation semantics.
+
+There is no universally optimal value. Resolution, frame count, GPU, kernel
+version, and checkpoint all affect both quality and latency. Start from the
+logged `num_blocks`, test several keep ratios on the target workload, and
+compare output quality against the same checkpoint running dense attention.
+
+## Checkpoint behavior
+
+The backend does not expose a user-selectable gate mode. Wan checkpoints that
+contain learned `to_gate_compress` weights use that projection automatically.
+When those weights are absent, the unused projection is removed and VSA uses
+the sparse branch without learned compensation.
+
+FastVideo DMD checkpoints use their fixed distilled timestep schedule. Native
+Wan checkpoints keep their normal scheduler and inference-step configuration;
+selecting VSA does not turn a native checkpoint into a distilled model.
+
+## Verify routing and fallback
+
+Check the startup and first-forward logs instead of assuming that selecting
+the backend guarantees sparse execution:
+
+- `route=VSA` means top-k block selection is active.
+- `route=VSA_ALL_BLOCKS` means the FastVideo DMD checkpoint retained all
+  blocks through the VSA kernel.
+- `route=SDPA` or `FASTVIDEO_VSA falling back to SDPA: ...` means dense SDPA
+  executed; the warning includes the reason.
+
+The current kernel requires CUDA tensors in FP16 or BF16, 256-token blocks,
+standard `head_size**-0.5` scaling, equal Q/K/V head counts, no attention mask,
+and no active sequence-parallel context. NPU and XPU paths do not execute the
+FastVideo VSA CUDA kernel.

docs/user_guide/diffusion/attention_backends/huggingface_hub.md

@@ -0,0 +1,32 @@
+# Hugging Face Hub Attention Backends
+
+Hub backends load the same published kernels commonly used with Hugging Face
+Diffusers. They are useful when training and serving must use matching kernel
+implementations to reduce numerical drift and sampling divergence.
+
+| Backend | Kernel | Platform |
+| --- | --- | --- |
+| `FLASH_ATTN_HUB` | `kernels-community/flash-attn2` | Compatible CUDA GPUs |
+| `FLASH_ATTN_3_HUB` | `kernels-community/flash-attn3` | Hopper sm_90 or newer |
+
+## Installation
+
+```bash
+pip install kernels==0.14.1
+```
+
+If `kernels` is unavailable, vLLM-Omni warns and falls back to the
+corresponding local FlashAttention implementation. On CUDA devices older than
+Hopper, `FLASH_ATTN_3_HUB` falls back to `FLASH_ATTN_HUB`.
+
+## Usage
+
+```bash
+export DIFFUSION_ATTENTION_BACKEND=FLASH_ATTN_HUB
+
+# Equivalent CLI selection
+vllm-omni serve <model> --diffusion-attention-backend FLASH_ATTN_HUB
+```
+
+For per-role kernel alignment, configure the Hub backend on only the relevant
+role; see the [attention backend overview](../attention_backends.md#per-role-configuration).

docs/user_guide/diffusion/attention_backends/rainfusion.md

@@ -0,0 +1,72 @@
+# RainFusion Attention
+
+`RAINFUSION_ATTN` runs MindIE-SD RainFusion (`rf_v2`) block-sparse video
+attention on Ascend NPU. It pools 128-token key blocks, ranks them per query
+block, and attends to the highest-scoring blocks after arranging video tokens
+in `(t, h, w)` order.
+
+Only the video segment is sparse. Prefix rows such as text, visual conditions,
+and audio, plus first-frame blocks, remain dense. Unsupported calls—including
+warmup steps, skipped layers, missing video geometry, and video segments below
+32 blocks—delegate to `FLASH_ATTN`, so compatible models can select RainFusion
+globally.
+
+## Configuration
+
+| Key | Valid values | Meaning |
+| --- | --- | --- |
+| `sparsity` | finite, `[0, 1]` | Nominal dropped-key-block fraction; default `0.8`; `0` disables sparsity |
+| `start_step` | integer, `>= 0` | Number of early denoise steps kept dense |
+| `skip_layers` | selector such as `"0-3,38"` | DiT blocks kept dense |
+
+```bash
+vllm-omni serve MiniMaxAI/MiniMax-H3 \
+  --diffusion-attention-config '{"default":{"backend":"RAINFUSION_ATTN",\
+    "block_sparse":{"sparsity":0.8,"start_step":0}}}'
+```
+
+```python
+from vllm_omni.diffusion.data import AttentionConfig, AttentionSpec, BlockSparseSpec
+
+config = AttentionConfig(
+    default=AttentionSpec(
+        backend="RAINFUSION_ATTN",
+        block_sparse=BlockSparseSpec(
+            sparsity=0.8,
+            start_step=0,
+            skip_layers="0-1",
+        ),
+    ),
+)
+```
+
+Tune `start_step`, then `sparsity`, then `skip_layers`. Increasing the dense
+early-step window is usually the cheapest way to recover global structure;
+use layer exclusions only after a same-seed dense comparison identifies
+sensitive blocks.
+
+## Requirements and compatibility
+
+RainFusion requires Ascend NPU and `mindiesd`. Selecting it on another
+platform raises. It is incompatible with ring sequence parallelism because
+the kernel needs the complete key sequence for ranking; use Ulysses sequence
+parallelism with `ring_degree=1`.
+
+## Geometry handling
+
+RainFusion handles arbitrary video grids by rearranging video spatially and,
+when necessary, promoting a real-video suffix to the always-kept prefix. It
+does not pad because `rf_v2` does not consume a padding attention mask.
+
+For MiniMax-H3 at 1344x768, grid `(62, 24, 42)` produces 62,496 video rows.
+The implementation promotes 2,976 rows and leaves 59,520 sparse rows
+(`465 x 128`) so the mask and kernel tiling remain aligned.
+
+Spatial grids aligned to 8x8 are still preferable. A latent height or width
+not divisible by 8 creates an always-kept suffix and reduces realized
+sparsity. At the image level, that means width and height divisible by 256.
+Protected video tails require a compatible MindIE-SD release.
+
+For common configuration and selector behavior, see the
+[attention backend overview](../attention_backends.md) and the
+[backend selection design](../../../design/feature/attention_backend_selection.md).

docs/user_guide/diffusion/attention_backends/sage.md

@@ -0,0 +1,63 @@
+# SageAttention
+
+SageAttention backends provide lossy low-precision attention for diffusion
+models. Validate output quality against `TORCH_SDPA` at the same seed before
+using either backend in production.
+
+## `SAGE_ATTN`
+
+`SAGE_ATTN` uses SageAttention 2.2 with INT8-quantized attention and FP16
+accumulation.
+
+### Installation
+
+Install SageAttention into the same environment as vLLM-Omni:
+
+```bash
+git clone https://github.com/thu-ml/SageAttention.git
+cd SageAttention
+export EXT_PARALLEL=4 NVCC_APPEND_FLAGS="--threads 8" MAX_JOBS=32
+pip install . --no-build-isolation
+```
+
+Verify the installation:
+
+```bash
+python -c "import sageattention; print(sageattention.__file__)"
+```
+
+Select it globally:
+
+```bash
+vllm-omni serve <model> --diffusion-attention-backend SAGE_ATTN
+```
+
+## `SAGE_ATTN_3`
+
+`SAGE_ATTN_3` uses the SageAttention3 Blackwell implementation.
+
+### SageAttention3 installation
+
+```bash
+git clone https://github.com/thu-ml/SageAttention.git
+cd SageAttention/sageattention3_blackwell
+python setup.py install
+```
+
+Verify the installation:
+
+```bash
+python -c "import sageattn3; print(sageattn3.__file__)"
+```
+
+```bash
+vllm-omni serve <model> --diffusion-attention-backend SAGE_ATTN_3
+```
+
+`SAGE_ATTN_3` requires CUDA, an importable `sageattn3`, and a Blackwell-class
+GPU. Its kernel assumes the query-head count equals the key/value-head count.
+GQA and MQA diffusion calls therefore fall back to PyTorch SDPA for
+correctness.
+
+For common configuration and platform routing, see the
+[attention backend overview](../attention_backends.md).

docs/user_guide/diffusion/attention_backends/trtllm.md

@@ -0,0 +1,103 @@
+# TRTLLM Attention
+
+`TRTLLM_ATTN` runs FlashInfer's trtllm-gen FMHA kernels. It supports dense
+BF16 attention plus optional Skip-Softmax sparsity or SAGE quantization.
+
+For common selection and per-role configuration, see the
+[attention backend overview](../attention_backends.md).
+
+## Requirements
+
+Selecting `TRTLLM_ATTN` requires:
+
+- datacenter Blackwell, sm_100 or sm_103;
+- `head_dim=128`;
+- FlashInfer; and
+- a model-declared packed or mask-free compatible path.
+
+It is not supported on workstation Blackwell (sm_120/sm_121). An incompatible
+explicit selection raises instead of silently changing the backend.
+
+## Dense mode
+
+Select `TRTLLM_ATTN` without `skip_softmax` or `quant` to run dense BF16:
+
+```bash
+vllm-omni serve <model> --diffusion-attention-backend TRTLLM_ATTN
+```
+
+## Skip-Softmax
+
+Skip-Softmax trades some fidelity for speed by skipping low-contribution
+attention work. See the [feature design](../../../design/feature/skip_softmax.md)
+for the algorithm and implementation contract.
+
+| Key | Valid values | Meaning |
+| --- | --- | --- |
+| `target_sparsity` | finite, `[0, 1]` | Point on a model-calibrated curve |
+| `threshold` | finite, `>= 0` | Direct threshold; mutually exclusive with `target_sparsity` |
+| `disabled_until_timestep` | finite, `[0, 1]` | Keep early high-noise steps dense until normalized `t <= D` |
+
+```bash
+vllm-omni serve Wan-AI/Wan2.2-T2V-A14B-Diffusers \
+  --diffusion-attention-config '{"default":{"backend":"TRTLLM_ATTN",\
+    "skip_softmax":{"target_sparsity":0.65,"disabled_until_timestep":0.86}}}'
+```
+
+```python
+from vllm_omni.diffusion.data import AttentionConfig, AttentionSpec, SkipSoftmaxSpec
+
+config = AttentionConfig(
+    default=AttentionSpec(
+        backend="TRTLLM_ATTN",
+        skip_softmax=SkipSoftmaxSpec(
+            target_sparsity=0.65,
+            disabled_until_timestep=0.86,
+        ),
+    ),
+)
+```
+
+Start with `target_sparsity=0.65` and `disabled_until_timestep=0.86`. Increase
+sparsity only after comparing quality against dense output at the same seed.
+The benefit grows with sequence length because only attention is accelerated.
+
+## SAGE quantization
+
+TRTLLM SAGE quantizes Q/K per block and V per channel so both attention
+matrix multiplications use low precision.
+
+| Key | Valid values | Meaning |
+| --- | --- | --- |
+| `dtype_qk` | `int8`, `fp8_e4m3` | Q/K quantization dtype; absent means dense |
+| `q_block_size` | `1`, `4`, `16` | Q scale block size; default `1` |
+| `k_block_size` | `1`, `4`, `16` | K scale block size; default `16` |
+
+V is always quantized per channel to FP8 E4M3 and K-smoothing is internal.
+Only block sizes with compiled kernels are accepted.
+
+```bash
+vllm-omni serve Wan-AI/Wan2.2-T2V-A14B-Diffusers \
+  --diffusion-attention-config '{"default":{"backend":"TRTLLM_ATTN",\
+    "quant":{"dtype_qk":"fp8_e4m3","q_block_size":1,"k_block_size":16}}}'
+```
+
+```python
+from vllm_omni.diffusion.data import AttentionConfig, AttentionSpec, AttnQuantSpec
+
+config = AttentionConfig(
+    default=AttentionSpec(
+        backend="TRTLLM_ATTN",
+        quant=AttnQuantSpec(
+            dtype_qk="fp8_e4m3",
+            q_block_size=1,
+            k_block_size=16,
+        ),
+    ),
+)
+```
+
+This path requires FlashInfer 0.6.16rc1 or newer. FP8 Q/K kernels are
+available on sm_100 and sm_103; INT8 Q/K kernels are compiled for sm_100 only.
+The shared `AttnQuantSpec` is backend typed: values intended for
+`FLASHINFER_ATTN` are rejected here and vice versa.

docs/user_guide/diffusion/cpu_offload.md

@@ -1,389 +1,74 @@
 # CPU Offloading for Diffusion Models
 
-## Overview
+vLLM-Omni provides three CPU-offload strategies for diffusion models. Choose
+the coarsest strategy that meets your memory target: finer-grained streaming
+saves more device memory but adds more transfer and setup overhead.
 
-vLLM-Omni provides two offloading strategies to reduce GPU memory usage for diffusion models:
+For the shared factory, discovery, and lifecycle contract, see the
+[CPU Offloading design](../../design/feature/offloader/README.md).
 
-1. **Model-level (Sequential) Offloading**: Mutual exclusion between DiT model and encoder - only one is on GPU at a time.
-2. **Layerwise (Blockwise) Offloading**: Keeps only one transformer block on GPU at a time with compute-memory overlap.
+## Choose a strategy
 
-Both strategies use pinned memory for faster CPU-GPU transfers. The strategies are **mutually exclusive** for now - if both are enabled, layerwise takes priority.
+| Strategy | Device residency | Parallel scope | Guide |
+| --- | --- | --- | --- |
+| Model-level (sequential) | One pipeline component group at a time | Single device | [Model-Level Offloading](offloader/cpu_offload.md) |
+| Layerwise (blockwise) | One transformer block, with next-block prefetch | Single device | [Layerwise Offloading](offloader/layerwise_offload.md) |
+| Distributed layerwise | Fixed two-block device buffer; optional host-weight sharding and AllGather | Multiple GPU/NPU ranks | [Distributed Layerwise Offloading](offloader/distributed_layerwise_offload.md) |
 
+All strategies use pinned host memory for faster transfers where applicable.
+Configuration priority is:
+
+1. Distributed layerwise offloading.
+2. Layerwise offloading.
+3. Model-level offloading.
+
+Treat the flags as mutually exclusive. If more than one is enabled, the
+higher-priority strategy is selected.
+
+## Quick selection
+
+- Use [model-level offloading](offloader/cpu_offload.md) when swapping whole
+  encoders and DiTs is enough to fit and phase-boundary transfers are
+  acceptable.
+- Use [layerwise offloading](offloader/layerwise_offload.md) for compute-heavy video
+  DiTs where block transfers can overlap computation.
+- Use [distributed layerwise offloading](offloader/distributed_layerwise_offload.md)
+  when a multi-rank deployment also needs bounded device residency and,
+  optionally, sharded host weights.
+
+## Supported models
+
+| Architecture | Example models | DiT class | Model-level | Layerwise | Distributed layerwise | Layerwise block attributes |
+| --- | --- | --- | :---: | :---: | :---: | --- |
+| Flux2Pipeline | `black-forest-labs/FLUX.2-dev` | `Flux2Transformer2DModel` | yes | yes | — | `transformer_blocks`, `single_transformer_blocks` |
+| LongCatImagePipeline | `meituan-longcat/LongCat-Image` | `LongCatImageTransformer2DModel` | — | yes | — | `transformer_blocks`, `single_transformer_blocks` |
+| NextStep11Pipeline | `stepfun-ai/NextStep-1.1` | `NextStepModel` | — | yes | — | `layers` |
+| OvisImagePipeline | `AIDC-AI/Ovis-Image-7B` | `OvisImageTransformer2DModel` | — | yes | — | `transformer` |
+| QwenImagePipeline | `Qwen/Qwen-Image` | `QwenImageTransformer2DModel` | yes | yes | — | `transformer_blocks` |
+| StableDiffusionXLPipeline | `stabilityai/stable-diffusion-xl-base-1.0` | `SDXLUNet2DConditionModel` | yes | yes | — | `down_blocks`, `up_blocks` |
+| StableDiffusion3Pipeline | `stabilityai/stable-diffusion-3.5-medium` | `SD3Transformer2DModel` | — | yes | — | `transformer_blocks` |
+| Wan22I2VPipeline | `Wan-AI/Wan2.2-I2V-A14B-Diffusers` | `WanTransformer3DModel` | yes | yes | — | `blocks` |
+| Wan22Pipeline | `Wan-AI/Wan2.2-T2V-A14B-Diffusers` | `WanTransformer3DModel` | yes | yes | — | `blocks` |
+| BagelPipeline | `ByteDance-Seed/BAGEL-7B-MoT` | `Qwen2MoTModel` | — | yes | — | `layers`, customized modules |
+| Cosmos3OmniDiffusersPipeline | `nvidia/Cosmos3-Nano`, `nvidia/Cosmos3-Super` | `Cosmos3VFMTransformer`, `Cosmos3LanguageModel` | yes | yes | yes | `layers`, `gen_layers` |
+
+Model-level support requires discoverable DiT and encoder components.
+Layerwise support requires transformer block topology. Distributed support
+reuses that topology but still requires validation for the model, checkpoint,
+and parallel configuration.
+
+## Compatibility anchors
+
+These headings preserve links to sections that moved into dedicated guides.
 
 ## Model-level (Sequential) Offloading
 
-### How It Works
-
-Model-level offloading implements mutual exclusion between DiT transformer and encoder modules using pre forward hooks:
-
-- **When encoders run**: DiT transformer is offloaded to CPU
-- **When DiT runs**: Encoders are offloaded to CPU, if more than one dit models, only one loaded on GPU, others get offloaded to CPU.
-- **VAE**: Stays resident on GPU
-
-Before each module's forward pass, the hook automatically moves it to GPU while offloading the other module group to CPU. Transfers use pinned memory for speed.
-
-### Usage
-
-**Python API:**
-```python
-from vllm_omni import Omni
-
-m = Omni(model="Wan-AI/Wan2.2-T2V-A14B-Diffusers", enable_cpu_offload=True)
-```
-
-**CLI:**
-```bash
-vllm serve Wan-AI/Wan2.2-T2V-A14B-Diffusers --omni --enable-cpu-offload
-```
-
-### To Support a Model
-
-Implement the `SupportsComponentDiscovery` protocol to declare which
-submodules serve as pipeline components (used by offloading, HSDP
-sharding, and other framework features):
-
-```python
-from typing import ClassVar
-from vllm_omni.diffusion.models.interface import SupportsComponentDiscovery
-
-class MyPipeline(nn.Module, SupportsComponentDiscovery):
-    _dit_modules: ClassVar[list[str]] = ["transformer"]
-    _encoder_modules: ClassVar[list[str]] = ["text_encoder", "vision_model"]
-    _vae_modules: ClassVar[list[str]] = ["vae"]
-    _resident_modules: ClassVar[list[str]] = []  # optional
-
-    def __init__(self):
-        super().__init__()
-        self.transformer = ...     # DiT — stays on GPU during denoising
-        self.text_encoder = ...    # Encoder — offloaded to CPU during denoising
-        self.vision_model = ...    # Encoder — offloaded to CPU during denoising
-        self.vae = ...             # VAE — always on GPU
-```
-
-- `_dit_modules`: attribute names of denoising submodules (kept on GPU
-  during the diffusion loop).
-- `_encoder_modules`: attribute names of encoder/vision submodules
-  (offloaded to CPU during the diffusion loop).
-- `_vae_modules`: attribute names of VAE(s) (always kept on GPU, not
-  part of the mutual exclusion hooks).
-- `_resident_modules`: attribute names of small submodules that must
-  stay on GPU during layerwise offloading (e.g. embedders, connectors).
-  Optional — defaults to `[]`.
-
-All attribute names support dotted paths for nested submodules
-(e.g. `"pipe.transformer"`, `"bagel.time_embedder"`).
-
-Both DiT and encoder lists are needed because the offload hooks use
-mutual exclusion: when one group runs, the other moves to CPU.
-
-### Limitations
-- Cold start latency increases
-- Adds overhead from CPU-GPU transfers between encoder and denoising phases
-- Support single GPU only for now
-
-
-### Component offloading for split models (e.g. Cosmos3)
-
-Some models split their transformer into mutually-exclusive *components* that run
-in different phases of a single forward pass rather than as separate pipeline
-components -- e.g. Cosmos3's understanding (reasoner) component runs once per
-generation while the generation (generator) component runs every denoising step.
-Such models have no separate text encoder to swap against, so the transformer
-owns a small model-local offload path and wraps each phase with
-`with self._offload_context(name):`
-
-```python
-class Cosmos3VFMTransformer(nn.Module):
-    def forward(self, ...):
-        with self._offload_context("reasoner"):
-            ...  # understanding pass, runs once
-        with self._offload_context("generator"):
-            ...  # denoising pass, runs every step
-```
-
-Model-level offloading then keeps exactly one component GPU-resident at a time
-(the other on CPU), reusing the same `SequentialOffloadHook` `.to()` movers. The
-pipeline opts in by exposing `enable_omni_model_cpu_offload` (which drives the
-transformer's `enable_model_cpu_offload` and pins the VAE). Layerwise offloading
-works for these models too -- each component declares its own block container via
-`_layerwise_offload_blocks_attrs`.
-
+See [Model-Level Offloading](offloader/cpu_offload.md).
 
 ## Layerwise (Blockwise) Offloading
 
-### How It Works
-
-Layerwise offloading keeps only one transformer block on GPU at a time.
-
-As each block completes, the next block is prefetched to GPU while the current block is freed. The pre and forward hooks utilized by layerwise offloading apply a separate CUDA stream (`copy_stream`) to overlap weight transfer with computation, and retain flattened tensors in pinned CPU memory for block parameters re-materialization. Encoders, VAE, and non-block DiT modules (embeddings, norms) always stay on GPU.
-
-**Execution Flow:**
-
-| Block | Pre-forward Hook | Forward | Post-forward Hook |
-|-------|------------------|---------|-------------------|
-| block-0 | Prefetch block-1 (async) | Compute block-0 | Free block-0 |
-| block-1 | Prefetch block-2 (async) | Compute block-1 | Free block-1 |
-| ... | ... | ... | ... |
-| block-(n-1) | **Prefetch block-0** (async) | Compute block-(n-1) | Free block-(n-1) |
-
-Each transformer block has a `LayerwiseOffloadHook` that prefetches the next block before forward and frees the current block after forward.
-
-Layerwise offloading is primarily recommended for large **video generation models** where the compute cost per block is high enough to effectively overlap with memory prefetch operations. For example, Wan2.2 T2V and I2V pipelines.
-
-### Usage
-
-**Python API:**
-```python
-from vllm_omni import Omni
-
-# Text-to-video
-m = Omni(model="Wan-AI/Wan2.2-T2V-A14B-Diffusers", enable_layerwise_offload=True)
-
-# Or image-to-video
-m = Omni(model="Wan-AI/Wan2.2-I2V-A14B-Diffusers", enable_layerwise_offload=True)
-```
-
-**CLI:**
-```bash
-# Text-to-video
-vllm serve Wan-AI/Wan2.2-T2V-A14B-Diffusers --omni --enable-layerwise-offload
-
-# Or image-to-video
-vllm serve Wan-AI/Wan2.2-I2V-A14B-Diffusers --omni --enable-layerwise-offload
-```
-
-### To Support a Model
-
-Models must define the blocks attribute name for layerwise offloading:
-
-```python
-class WanTransformer3DModel(nn.Module):
-    _layerwise_offload_blocks_attrs = ["blocks"]  # Attribute names containing transformer blocks
-
-    def __init__(self):
-        self.blocks = nn.ModuleList([...])  # Transformer blocks
-```
-
-For models with multiple block types:
-
-```python
-class Flux2Transformer2DModel(nn.Module):
-    _layerwise_offload_blocks_attrs = ["transformer_blocks", "single_transformer_blocks"]
-```
-
-### Limitations
-- Cold start latency increases because offloaded components must be moved to CPU
-  during setup; layerwise offload may add extra weight consolidation and pinning
-  work.
-- Performance depends on compute cost and H2D bandwidth as well
-- Support single GPU only for now
-
+See [Layerwise Offloading](offloader/layerwise_offload.md).
 
 ## Distributed Layerwise Offloading
 
-### How It Works
-
-Distributed layerwise offloading extends single-GPU layerwise offloading to
-multi-device deployments. In the default mode, each rank stores only
-**1/dp_size** of the model weights in host memory; full layer weights are
-reconstructed at runtime via **AllGather** on a dedicated communication stream,
-overlapped with computation via a fixed double-buffer scheme. The
-`--dlo-no-use-allgather` mode instead streams standard-loader rank-local
-weights independently and does not shard weights across ranks.
-
-For the implementation design and compatibility matrix, see
-[Distributed Layerwise Offload](../../design/feature/distributed_layerwise_offload.md).
-
-**Key features:**
-- **Weight sharding + AllGather**: each rank stores 1/dp_size of weights,
-  reconstructed per-layer via `all_gather_into_tensor`
-- **Fixed double-buffer**: exactly 2 transformer blocks on each device at any
-  time, regardless of model size
-- **DP multi-concurrency**: N concurrent requests processed in parallel
-  (AllGather only gathers weight shards, which are request-independent)
-- **mmap weight loading**: weights loaded as mmap views pointing to shared OS
-  page cache, eliminating O(dp_size × model_size) RSS during model creation
-- **Platform-agnostic**: works on NVIDIA GPU (CUDA/NCCL) and Ascend NPU
-  (CANN/HCCL) via vLLM-Omni's platform abstraction
-
-**Execution Flow (per device):**
-
-```
-Compute Stream:  [Layer N]          [Layer N+1]          [Layer N+2]
-H2D Stream:      [H2D Shard N+1]   [H2D Shard N+2]
-AllGather:       [AG N+1]          [AG N+2]
-Slot Usage:      Slot0: Layer N    Slot1: Layer N+1
-```
-
-### Usage
-
-**CLI:**
-```bash
-# 4× GPU/NPU with AllGather (recommended)
-vllm serve /path/to/model --omni \
-  --enable-distributed-layerwise-offload \
-  --data-parallel-size 4
-
-# Without DLO AllGather (each rank streams standard-loader rank-local weights)
-vllm serve /path/to/model --omni \
-  --enable-distributed-layerwise-offload \
-  --data-parallel-size 4 \
-  --dlo-no-use-allgather
-
-# With SP instead of DP (long sequences)
-vllm serve /path/to/model --omni \
-  --enable-distributed-layerwise-offload \
-  --usp 4
-```
-
-**Python API:**
-```python
-from vllm_omni import Omni
-
-m = Omni(
-    model="/path/to/model",
-    enable_distributed_layerwise_offload=True,
-    dlo_use_allgather=True,  # default
-)
-```
-
-### CLI Flags
-
-| Flag | Description | Default |
-|------|-------------|---------|
-| `--enable-distributed-layerwise-offload` | Enable DLO | `false` |
-| `--data-parallel-size N` | Number of DP ranks; DLO weight-sharding group when AllGather is enabled | `1` |
-| `--dlo-use-allgather` | Shard + AllGather (saves CPU, requires concurrent requests) | `true` |
-| `--dlo-no-use-allgather` | Use standard-loader rank-local weights; no DLO sharding or AllGather | `false` |
-
-### How model weights are loaded (mmap path)
-
-When DLO + AllGather is active, the offloader:
-
-1. **Saves non-persistent buffers** (e.g. RoPE `inv_freq`, timestep `freqs`)
-   from the normally-created transformer
-2. **Converts to meta device** via `to_empty(device="meta")`, releasing random
-   initialization weights
-3. **Loads checkpoint weights as mmap views** via `safe_open().get_tensor()`,
-   which return views into the OS page cache (shared across all ranks, 0 RSS)
-4. **Calls `post_load_weights()`** to apply model-specific dtype conversions
-   (e.g. Cosmos3's `time_embedder` → FP32)
-5. **Restores non-persistent buffers** from saved copies
-
-This approach requires **zero model-specific code changes** — no pipeline or
-transformer modifications are needed.
-
-### OffloadPlan (declarative topology metadata)
-
-Models can optionally declare an `OffloadPlan` class variable to provide
-topology metadata (block attribute names, submodules to offload) without
-any offload-specific logic:
-
-```python
-from vllm_omni.diffusion.offloader import OffloadPlan
-
-class MyPipeline(nn.Module):
-    _dit_modules = ["transformer"]
-    _offload_plan = OffloadPlan(
-        block_attrs={"transformer": ("blocks",)},
-        offload_submodules={"context_encoder": "layers"},
-    )
-```
-
-When not declared, the offloader falls back to `_layerwise_offload_blocks_attrs`
-and heuristic attribute search.  This is backward-compatible — existing models
-work without any changes.
-
-### DP Multi-concurrency
-
-When `--data-parallel-size > 1` and AllGather is enabled, the scheduler batches
-up to `dp_size` requests per denoise step.  Each DP rank processes a different
-request while AllGather synchronizes only weight shards (request-independent).
-
-**Requirements for concurrent requests:**
-- `num_inference_steps` must be specified explicitly (None is not allowed)
-- All concurrent requests must have the same `num_inference_steps` value
-- These constraints exist because AllGather is a collective that requires
-  every rank to participate at each denoise step
-
-### Limitations
-- Online quantization (FP8) is incompatible with mmap loading — falls back to
-  regular loading only when the model is outside the DLO+AllGather mmap path;
-  online quantization with DLO+AllGather is rejected. Use
-  `--dlo-no-use-allgather` or disable online quantization.
-- Tensor Parallel > 1 is rejected in the DLO+AllGather mmap path because TP-aware
-  weight-loader callbacks are bypassed. The no-AllGather path retains the
-  standard TP loader and is experimental.
-- HSDP + AllGather is rejected because it would double-shard parameters. HSDP
-  with no-AllGather is accepted at configuration level, but has limited
-  end-to-end validation.
-- `num_inference_steps=None` is not allowed in DP multi-concurrency mode
-
-**Module Discovery**
-
-The offloader discovers pipeline components in two ways:
-
-1. **Protocol-based** (preferred): If the pipeline implements
-    `SupportsComponentDiscovery`, its `_dit_modules`, `_encoder_modules`,
-    `_vae_modules`, and `_resident_modules` class variables are used
-    directly.  All attribute names support dotted paths (e.g.
-    `"pipe.transformer"`, `"bagel.time_embedder"`) for nested submodules.
-
-2. **Fallback attribute scan**: Otherwise, the offloader scans for
-    well-known attribute names:
-    - **DiT modules**: `transformer`, `transformer_2`, `dit`, `sr_dit`, `language_model`, `transformer_blocks`, `model`
-    - **Encoders**: `text_encoder`, `text_encoder_2`, `text_encoder_3`, `image_encoder`
-    - **VAE**: `vae`, `audio_vae`
-
-**Hook System**
-
-Both strategies use vLLM-Omni's hook registry system (`HookRegistry` and `ModelHook`) to register pre/post forward callbacks on modules, enabling automatic swapping without modifying model code.
-
-**Backend Architecture**
-
-```
-OffloadBackend (base class)
-├── ModelLevelOffloadBackend → uses SequentialOffloadHook (.to() swap)
-│                              (delegates to a pipeline's enable_omni_model_cpu_offload
-│                               for split models like Cosmos3)
-├── LayerWiseOffloadBackend → uses LayerwiseOffloadHook
-│                          (single-GPU, full weights on host)
-└── DistributedLayerwiseOffloadBackend → uses DistributedLayerwiseOffloadHook
-                                         (multi-GPU, sharded AllGather or rank-local mode)
-```
-
-Factory function `get_offload_backend()` selects the appropriate backend based on
-configuration.
-
-For split models, `ModelLevelOffloadBackend.enable()` detects a pipeline's
-`enable_omni_model_cpu_offload` hook and delegates to it; Cosmos3 then swaps its
-reasoner/generator components inside the model forward pass.
-
-
-## Supported Models
-
-| Architecture | Example Models | DiT Class | Model-Level Offload | Layerwise Offload | Distributed Layerwise Offload | Blocks Attrs (Layerwise specific) |
-|--------------|----------------|-----------|---------------------|-------------------|-------------------------------|-----------------------------------|
-| Flux2Pipeline | `black-forest-labs/FLUX.2-dev` | `Flux2Transformer2DModel` | ✓ | ✓ | - | `"transformer_blocks"`, `"single_transformer_blocks"` |
-| LongCatImagePipeline | `meituan-longcat/LongCat-Image` | `LongCatImageTransformer2DModel` | - | ✓ | - | `"transformer_blocks"`, `"single_transformer_blocks"` |
-| NextStep11Pipeline | `stepfun-ai/NextStep-1.1` | `NextStepModel` | - | ✓ | - | `"layers"` |
-| OvisImagePipeline | `AIDC-AI/Ovis-Image-7B` | `OvisImageTransformer2DModel` | - | ✓ | - | `"transformer"` |
-| QwenImagePipeline | `Qwen/Qwen-Image` | `QwenImageTransformer2DModel` | ✓ | ✓ | - | `"transformer_blocks"` |
-| StableDiffusionXLPipeline | `stabilityai/stable-diffusion-xl-base-1.0` | `SDXLUNet2DConditionModel` | ✓ | ✓ | - | `"down_blocks"`, `"up_blocks"` |
-| StableDiffusion3Pipeline | `stabilityai/stable-diffusion-3.5-medium` | `SD3Transformer2DModel` | - | ✓ | - | `"transformer_blocks"` |
-| Wan22I2VPipeline | `Wan-AI/Wan2.2-I2V-A14B-Diffusers` | `WanTransformer3DModel` | ✓ | ✓ | - | `"blocks"` |
-| Wan22Pipeline | `Wan-AI/Wan2.2-T2V-A14B-Diffusers` | `WanTransformer3DModel` | ✓ | ✓ | - | `"blocks"` |
-| SoulXSingerPipeline / SoulXSingerSVCPipeline | `Soul-AILab/SoulX-Singer` | `DiffLlama` (`cfm_decoder.model.diff_estimator`) | ✓ | ✓ | - | `"layers"` |
-| BagelPipeline | `ByteDance-Seed/BAGEL-7B-MoT` | `Qwen2MoTModel` | - | ✓ | - | `"layers"`, `"customized modules"` |
-| Cosmos3OmniDiffusersPipeline | `nvidia/Cosmos3-Nano`, `nvidia/Cosmos3-Super` | `Cosmos3VFMTransformer`, `Cosmos3LanguageModel` | ✓ | ✓ | ✓ | `"layers"`, `"gen_layers"` |
-
-**Notes:**
-- Model-Level Offloading is expected to be supported by all common diffusion models (DiT and encoders) naturally
-- Layerwise Offloading requires DiT class to define `_layerwise_offload_blocks_attrs` pointing to transformer blocks
-- Distributed Layerwise Offloading reuses the layerwise block topology, but the
-  model and parallelism combination still needs validation. AllGather mode
-  additionally requires a compatible checkpoint/mmap layout; no-AllGather mode
-  retains the standard loader's rank-local tensors. See the
-  [DLO design](../../design/feature/distributed_layerwise_offload.md) and
-  [Cosmos3 DistOffload recipe](https://github.com/vllm-project/vllm-omni/blob/main/recipes/cosmos3/Cosmos3-DistOffload.md)
-  for usage examples.
+See [Distributed Layerwise Offloading](offloader/distributed_layerwise_offload.md).

docs/user_guide/diffusion/execution_modes.md

@@ -101,8 +101,16 @@ set `DIFFUSION_ATTENTION_BACKEND=TORCH_SDPA` or configure
 `--max-num-seqs >1`. See the
 [HunyuanImage-3.0 recipe](https://github.com/vllm-project/vllm-omni/blob/main/recipes/Tencent/HunyuanImage-3.0-Instruct.md)
 for its validated configuration. Helios supports single-request step execution only: use
-`--step-execution --max-num-seqs 1` for Helios. Consult the selected pipeline's
-documentation and source for the latest support status.
+`--step-execution --max-num-seqs 1` for Helios. MiniMax H3 supports step-wise
+continuous batching by packing co-batched requests into one sequence that keeps
+a separate attention document per request; that layout needs a backend which
+honors the packed `cu_seqlens` metadata, so run it with
+`--diffusion-attention-backend FLASH_ATTN` (other backends stay correct but
+fall back to one transformer forward per request). Batching H3 does not improve
+its throughput — see the measured numbers in the
+[MiniMax-H3 recipe](https://github.com/vllm-project/vllm-omni/blob/main/recipes/MiniMaxAI/MiniMax-H3.md)
+— so keep `--max-num-seqs 1` unless you need step-level scheduling. Consult the
+selected pipeline's documentation and source for the latest support status.
 
 ## Streaming Output
 
@@ -207,15 +215,13 @@ outputs = omni.generate(
 )
 ```
 
-## Stage Configuration
+## Deploy Configuration
 
 Use the equivalent engine arguments in a deployment YAML:
 
 ```yaml
-stage_args:
+stages:
   - stage_id: 0
-    stage_type: diffusion
-    engine_args:
     step_execution: false
     max_num_seqs: 4
     request_batch_max_wait_ms: 20