diff --git a/examples/online_serving/image_to_video/README.md b/examples/online_serving/image_to_video/README.md
index 8f5b65061..3ae3a5291 100644
--- a/examples/online_serving/image_to_video/README.md
+++ b/examples/online_serving/image_to_video/README.md
@@ -1,9 +1,9 @@
 # Image-To-Video
 
-This shared example demonstrates online image-to-video generation with
-vLLM-Omni. The startup script defaults to Wan2.2, but `MODEL` can select any
-supported image-to-video model. The client sends the canonical Videos API
-request; the server side owns model-specific prompt processing.
+This example demonstrates online image-to-video generation with vLLM-Omni.
+The startup script defaults to Wan2.2, but `MODEL` can select any supported
+image-to-video model. The existing curl helper remains a concrete Wan2.2
+request, while the server side owns model-specific prompt processing.
 
 ## Start Server
 
diff --git a/examples/online_serving/image_to_video/run_curl_image_to_video.sh b/examples/online_serving/image_to_video/run_curl_image_to_video.sh
index 40ac9bc6b..6f6a6f96d 100644
--- a/examples/online_serving/image_to_video/run_curl_image_to_video.sh
+++ b/examples/online_serving/image_to_video/run_curl_image_to_video.sh
@@ -1,24 +1,14 @@
 #!/bin/bash
-# Shared image-to-video curl example using the async video job API.
+# Wan2.2 image-to-video curl example using the async video job API.
 
 set -euo pipefail
 
 INPUT_IMAGE="${INPUT_IMAGE:-../../offline_inference/image_to_video/qwen-bear.png}"
 BASE_URL="${BASE_URL:-http://localhost:8099}"
-OUTPUT_PATH="${OUTPUT_PATH:-image_to_video_output.mp4}"
-PROMPT="${PROMPT:-A bear playing with yarn, smooth motion}"
+OUTPUT_PATH="${OUTPUT_PATH:-wan22_i2v_output.mp4}"
 NEGATIVE_PROMPT="${NEGATIVE_PROMPT:-}"
 SAMPLE_SOLVER="${SAMPLE_SOLVER:-}"
 POLL_INTERVAL="${POLL_INTERVAL:-2}"
-VIDEO_SECONDS="${VIDEO_SECONDS:-}"
-SIZE="${SIZE:-}"
-FPS="${FPS:-}"
-NUM_INFERENCE_STEPS="${NUM_INFERENCE_STEPS:-}"
-GUIDANCE_SCALE="${GUIDANCE_SCALE:-}"
-GUIDANCE_SCALE_2="${GUIDANCE_SCALE_2:-}"
-BOUNDARY_RATIO="${BOUNDARY_RATIO:-}"
-FLOW_SHIFT="${FLOW_SHIFT:-}"
-SEED="${SEED:-}"
 
 if [ ! -f "$INPUT_IMAGE" ]; then
     echo "Input image not found: $INPUT_IMAGE"
@@ -28,28 +18,19 @@ fi
 create_cmd=(
   curl -sS -X POST "${BASE_URL}/v1/videos"
   -H "Accept: application/json"
-  -F "prompt=${PROMPT}"
+  -F "prompt=A bear playing with yarn, smooth motion"
   -F "input_reference=@${INPUT_IMAGE}"
+  -F "seconds=2"
+  -F "size=832x480"
+  -F "fps=16"
+  -F "num_inference_steps=40"
+  -F "guidance_scale=1.0"
+  -F "guidance_scale_2=1.0"
+  -F "boundary_ratio=0.875"
+  -F "flow_shift=12.0"
+  -F "seed=42"
 )
 
-append_optional_field() {
-  local name="$1"
-  local value="$2"
-  if [ -n "${value}" ]; then
-    create_cmd+=(-F "${name}=${value}")
-  fi
-}
-
-append_optional_field "seconds" "${VIDEO_SECONDS}"
-append_optional_field "size" "${SIZE}"
-append_optional_field "fps" "${FPS}"
-append_optional_field "num_inference_steps" "${NUM_INFERENCE_STEPS}"
-append_optional_field "guidance_scale" "${GUIDANCE_SCALE}"
-append_optional_field "guidance_scale_2" "${GUIDANCE_SCALE_2}"
-append_optional_field "boundary_ratio" "${BOUNDARY_RATIO}"
-append_optional_field "flow_shift" "${FLOW_SHIFT}"
-append_optional_field "seed" "${SEED}"
-
 if [ -n "${NEGATIVE_PROMPT}" ]; then
   create_cmd+=(-F "negative_prompt=${NEGATIVE_PROMPT}")
 fi
diff --git a/examples/online_serving/text_to_image/README.md b/examples/online_serving/text_to_image/README.md
index 2c9bc2cb6..a57a917d0 100644
--- a/examples/online_serving/text_to_image/README.md
+++ b/examples/online_serving/text_to_image/README.md
@@ -1,9 +1,9 @@
 # Text-To-Image
 
-This shared example demonstrates online text-to-image generation with vLLM-Omni.
-The startup script defaults to Qwen-Image, but `MODEL` can select any supported
-text-to-image model. Clients send the canonical Images API request; the server
-side owns model-specific prompt processing.
+This example demonstrates online text-to-image generation with vLLM-Omni. The
+startup script and Python clients default to Qwen-Image, but `MODEL` can select
+any supported text-to-image model. The existing curl helper remains a concrete
+Qwen-Image request, while the server side owns model-specific prompt processing.
 
 ## Start Server
 
diff --git a/examples/online_serving/text_to_image/run_curl_text_to_image.sh b/examples/online_serving/text_to_image/run_curl_text_to_image.sh
index a3961e66e..151df8948 100755
--- a/examples/online_serving/text_to_image/run_curl_text_to_image.sh
+++ b/examples/online_serving/text_to_image/run_curl_text_to_image.sh
@@ -1,21 +1,10 @@
 #!/bin/bash
-# Shared text-to-image curl example
+# Qwen-Image text-to-image curl example
 
-set -euo pipefail
-
-BASE_URL="${BASE_URL:-http://localhost:8091}"
-PROMPT="${PROMPT:-a dragon laying over the spine of the Green Mountains of Vermont}"
-SIZE="${SIZE:-}"
-SEED="${SEED:-}"
-OUTPUT_PATH="${OUTPUT_PATH:-text_to_image_output.png}"
-
-curl -sS -X POST "${BASE_URL}/v1/images/generations" \
+curl -X POST http://localhost:8091/v1/images/generations \
   -H "Content-Type: application/json" \
-  -d "$(jq -n --arg prompt "${PROMPT}" --arg size "${SIZE}" --arg seed "${SEED}" \
-    '{prompt: $prompt}
-     + (if $size == "" then {} else {size: $size} end)
-     + (if $seed == "" then {} else {seed: ($seed | tonumber)} end)')" \
-  | jq -r '.data[0].b64_json' \
-  | base64 -d > "${OUTPUT_PATH}"
-
-echo "Saved image to ${OUTPUT_PATH}"
+  -d '{
+    "prompt": "a dragon laying over the spine of the Green Mountains of Vermont",
+    "size": "1024x1024",
+    "seed": 42
+  }' | jq -r '.data[0].b64_json' | base64 -d > dragon.png