fix: handle ModelResponse cache return in responses bridge and drop dead acompletion check

cursoragent · yassin-berriai · cursoragent · commit bda3ee6f9df0 · 2026-05-18T12:16:48.000Z
Co-authored-by: Yassin Kortam &lt;yassin@berri.ai&gt;
diff --git a/litellm/caching/caching_handler.py b/litellm/caching/caching_handler.py
@@ -868,9 +868,7 @@ def _convert_cached_result_to_model_response(
         elif (call_type == "aresponses" or call_type == "responses") and isinstance(
             cached_result, dict
         ):
-            use_chat_completion_cache = kwargs.get(
-                "acompletion"
-            ) is True or _is_chat_completion_cached_dict(cached_result)
+            use_chat_completion_cache = _is_chat_completion_cached_dict(cached_result)
             if use_chat_completion_cache:
                 if kwargs.get("stream", False) is True:
                     bridge_call_type = (
diff --git a/litellm/completion_extras/litellm_responses_transformation/handler.py b/litellm/completion_extras/litellm_responses_transformation/handler.py
@@ -186,6 +186,8 @@ def completion(self, *args, **kwargs) -> Union[
             **request_data,
         )
 
+        from litellm.types.utils import ModelResponse
+
         stream = self._resolve_stream_flag(optional_params, litellm_params)
         if isinstance(result, ResponsesAPIResponse):
             return self.transformation_handler.transform_response(
@@ -201,6 +203,8 @@ def completion(self, *args, **kwargs) -> Union[
                 api_key=kwargs.get("api_key"),
                 json_mode=kwargs.get("json_mode"),
             )
+        elif isinstance(result, ModelResponse):
+            return result
         elif not stream:
             responses_api_response = self._collect_response_from_stream(result)
             return self.transformation_handler.transform_response(
@@ -269,6 +273,8 @@ async def acompletion(
             aresponses=True,
         )
 
+        from litellm.types.utils import ModelResponse
+
         stream = self._resolve_stream_flag(optional_params, litellm_params)
         if isinstance(result, ResponsesAPIResponse):
             return self.transformation_handler.transform_response(
@@ -284,6 +290,8 @@ async def acompletion(
                 api_key=kwargs.get("api_key"),
                 json_mode=kwargs.get("json_mode"),
             )
+        elif isinstance(result, ModelResponse):
+            return result
         elif not stream:
             responses_api_response = await self._collect_response_from_stream_async(
                 result
diff --git a/tests/local_testing/test_caching_handler.py b/tests/local_testing/test_caching_handler.py
@@ -1115,7 +1115,6 @@ def test_convert_cached_aresponses_bridge_chat_completion_stream():
         kwargs={
             "model": "gpt-5.4",
             "stream": True,
-            "acompletion": True,
             "messages": [{"role": "user", "content": "hi"}],
         },
         logging_obj=logging_obj,
