diff --git a/.buildkite/intel/pipeline-intel.yml b/.buildkite/intel/pipeline-intel.yml
index e68f59e53..acb135101 100644
--- a/.buildkite/intel/pipeline-intel.yml
+++ b/.buildkite/intel/pipeline-intel.yml
@@ -10,7 +10,7 @@ steps:
           DOCKER_BUILDKIT: "1"
           # Buildkite will automatically replace this with the actual commit hash
           VLLM_IMAGE_TAG: "${BUILDKITE_COMMIT}"
-          VLLM_VERSION: "v0.25.0"
+          VLLM_VERSION: "v0.26.0"
         priority: 100
         timeout_in_minutes: 60
         soft_fail: false
diff --git a/vllm_omni/platforms/interface.py b/vllm_omni/platforms/interface.py
index c4e4c6e37..9969849d9 100644
--- a/vllm_omni/platforms/interface.py
+++ b/vllm_omni/platforms/interface.py
@@ -109,6 +109,7 @@ class OmniPlatform(Platform):
         cls,
         selected_backend: str | None,
         head_size: int,
+        allow_trtllm_default: bool = False,
     ) -> str:
         """Get the diffusion attention backend class path for this platform.
 
@@ -119,6 +120,7 @@ class OmniPlatform(Platform):
             selected_backend: User-selected backend name (e.g., "FLASH_ATTN",
                 "TORCH_SDPA", "SAGE_ATTN"). If None, uses platform default.
             head_size: Attention head size.
+            allow_trtllm_default: Whether TRTLLM may be chosen as the default.
 
         Returns:
             Fully qualified class path of the selected backend.
diff --git a/vllm_omni/platforms/xpu/platform.py b/vllm_omni/platforms/xpu/platform.py
index 8c19868ed..26d3dccde 100644
--- a/vllm_omni/platforms/xpu/platform.py
+++ b/vllm_omni/platforms/xpu/platform.py
@@ -44,7 +44,9 @@ class XPUOmniPlatform(OmniPlatform, XPUPlatform):
         cls,
         selected_backend: str | None,
         head_size: int,
+        allow_trtllm_default: bool = False,
     ) -> str:
+        # XPU has no TRTLLM backend; arg accepted for signature parity but unused.
         compute_capability = torch.xpu.get_device_capability()
         # Intel Max 1100 and 1550 will not support flash_attn currently
         flash_attn_supported = compute_capability["architecture"] not in [13136561920]