[Frontend] Make model optional on all /derender request classes (#51463)

Signed-off-by: Vinay Damodaran <vrdn@hey.com>
Co-authored-by: Claude Opus 5 (1M context) <noreply@anthropic.com>

tests/entrypoints/scale_out/derender/test_derender.py

@@ -308,6 +308,25 @@ async def test_derender_chat_unknown_model(client):
     assert response.status_code == 404
 
 
+@pytest.mark.asyncio
+async def test_derender_chat_model_omitted_resolves_served_name(client):
+    """Omitting `model` resolves the served name rather than rejecting.
+
+    Mirrors test_serving_chat.py's "full name is returned when no model is
+    specified" assertion for the derender path. Asserts the resolved value,
+    not just the status, so the fallback is proven to have fired.
+    """
+    gen_req = await _render_chat(client)
+    synthetic_ids = gen_req["token_ids"][:3]
+
+    response = await client.post(
+        "/v1/chat/completions/derender",
+        json={"generate_response": _make_generate_response(synthetic_ids)},
+    )
+    assert response.status_code == 200
+    assert response.json()["model"] == MODEL_NAME
+
+
 # ---------------------------------------------------------------------------
 # Completion derender tests
 # ---------------------------------------------------------------------------

vllm/entrypoints/scale_out/derender/serving.py

@@ -166,16 +166,17 @@ class ServingDerender(BaseServing):
             total_tokens=prompt_tokens + completion_tokens,
         )
 
+        model_name = request.model or self.models.model_name()
         logger.debug(
             "derender_chat request_id=%s model=%s choices=%d completion_tokens=%d",
             gen.request_id,
-            request.model,
+            model_name,
             len(choices),
             completion_tokens,
         )
         return ChatCompletionResponse(
             id=gen.request_id,
-            model=request.model,
+            model=model_name,
             created=int(time.time()),
             choices=choices,
             usage=usage,
@@ -231,17 +232,18 @@ class ServingDerender(BaseServing):
             total_tokens=total_prompt_tokens + total_completion_tokens,
         )
 
+        model_name = request.model or self.models.model_name()
         logger.debug(
             "derender_completion request_id=%s model=%s choices=%d"
             " completion_tokens=%d",
             first.request_id,
-            request.model,
+            model_name,
             len(choices),
             total_completion_tokens,
         )
         return CompletionResponse(
             id=first.request_id,
-            model=request.model,
+            model=model_name,
             created=int(time.time()),
             choices=choices,
             usage=usage,
@@ -264,9 +266,10 @@ class ServingDerender(BaseServing):
         if error_check_ret is not None:
             return error_check_ret
 
+        model_name = request.model or self.models.model_name()
         try:
             chunk, updated_state = await self.online_derenderer.derender_chat_stream(
-                model=request.model,
+                model=model_name,
                 generate_chunk=request.generate_chunk,
                 state=request.stream_state,
                 chat_request=request.chat_request,
@@ -284,7 +287,7 @@ class ServingDerender(BaseServing):
         logger.debug(
             "derender_chat_stream request_id=%s model=%s delta_tokens=%d",
             request.generate_chunk.request_id,
-            request.model,
+            model_name,
             sum(
                 len(c.token_ids) for c in request.generate_chunk.choices if c.token_ids
             ),
@@ -304,12 +307,13 @@ class ServingDerender(BaseServing):
         if error_check_ret is not None:
             return error_check_ret
 
+        model_name = request.model or self.models.model_name()
         try:
             (
                 chunk,
                 updated_state,
             ) = await self.online_derenderer.derender_completion_stream(
-                model=request.model,
+                model=model_name,
                 generate_chunk=request.generate_chunk,
                 state=request.stream_state,
                 prompt_tokens=request.prompt_tokens,
@@ -325,7 +329,7 @@ class ServingDerender(BaseServing):
         logger.debug(
             "derender_completion_stream request_id=%s model=%s delta_tokens=%d",
             request.generate_chunk.request_id,
-            request.model,
+            model_name,
             sum(
                 len(c.token_ids) for c in request.generate_chunk.choices if c.token_ids
             ),

vllm/entrypoints/scale_out/token_in_token_out/protocol.py

@@ -277,8 +277,8 @@ class DerenderChatRequest(BaseModel):
     # --8<-- [start:derender-chat-request]
     stream: Literal[False] = False
 
-    model: str
-    """Served model name."""
+    model: str | None = None
+    """Served model name. Defaults to the server's served model name."""
 
     generate_response: GenerateResponse
     """The complete token-in / token-out engine response to derender."""
@@ -311,8 +311,8 @@ class DerenderCompletionRequest(BaseModel):
     # --8<-- [start:derender-completion-request]
     stream: Literal[False] = False
 
-    model: str
-    """Served model name."""
+    model: str | None = None
+    """Served model name. Defaults to the server's served model name."""
 
     generate_responses: list[GenerateResponse]
     """One response per prompt, parallel to the list[GenerateRequest]
@@ -428,7 +428,7 @@ class DerenderChatStreamRequest(BaseModel):
 
     stream: Literal[True]
 
-    model: str
+    model: str | None = None
     generate_chunk: GenerateStreamResponse
     """One SSE chunk from ``/inference/v1/generate`` (``stream=True``)."""
 
@@ -452,7 +452,7 @@ class DerenderCompletionStreamRequest(BaseModel):
 
     stream: Literal[True]
 
-    model: str
+    model: str | None = None
     generate_chunk: GenerateStreamResponse
     """One SSE chunk from ``/inference/v1/generate``."""