From 4e7115bc34818b991c38a0f9301ba8faf3785b5d Mon Sep 17 00:00:00 2001
From: Ishaan Jaff <ishaanjaffer0324@gmail.com>
Date: Tue, 1 Jul 2025 18:12:22 -0700
Subject: [PATCH] Bug Fix - responses api fix got multiple values for keyword
 argument 'litellm_trace_id'  (#12225)

* fix - handling trace id arg on responses api

* test_async_response_api_handler_merges_trace_id_without_error

* test_anthropic_with_responses_api
---
 litellm/proxy/proxy_config.yaml               |  5 ++-
 .../handler.py                                | 11 ++++--
 .../test_anthropic_responses_api.py           | 35 ++++++++++++++++++-
 .../test_e2e_openai_responses_api.py          |  9 +++++
 4 files changed, 54 insertions(+), 6 deletions(-)

diff --git a/litellm/proxy/proxy_config.yaml b/litellm/proxy/proxy_config.yaml
index 13724177c05..8a8fd6794e7 100644
--- a/litellm/proxy/proxy_config.yaml
+++ b/litellm/proxy/proxy_config.yaml
@@ -1,8 +1,7 @@
 model_list:
-  - model_name: groq/*
+  - model_name: anthropic/*
     litellm_params:
-      model: groq/*
-      api_key: bad
+      model: anthropic/*
   - model_name: openai/*
     litellm_params:
       model: openai/*
diff --git a/litellm/responses/litellm_completion_transformation/handler.py b/litellm/responses/litellm_completion_transformation/handler.py
index f960092f2bf..7dc182747df 100644
--- a/litellm/responses/litellm_completion_transformation/handler.py
+++ b/litellm/responses/litellm_completion_transformation/handler.py
@@ -56,6 +56,10 @@ class LiteLLMCompletionTransformationHandler:
                 responses_api_request=responses_api_request,
                 **kwargs,
             )
+        
+        completion_args = {}
+        completion_args.update(kwargs)
+        completion_args.update(litellm_completion_request)
 
         litellm_completion_response: Union[
             ModelResponse, litellm.CustomStreamWrapper
@@ -98,12 +102,15 @@ class LiteLLMCompletionTransformationHandler:
                 previous_response_id=previous_response_id,
                 litellm_completion_request=litellm_completion_request,
             )
+        
+        acompletion_args = {}
+        acompletion_args.update(kwargs)
+        acompletion_args.update(litellm_completion_request)
 
         litellm_completion_response: Union[
             ModelResponse, litellm.CustomStreamWrapper
         ] = await litellm.acompletion(
-            **litellm_completion_request,
-            **kwargs,
+            **acompletion_args,
         )
 
         if isinstance(litellm_completion_response, ModelResponse):
diff --git a/tests/llm_responses_api_testing/test_anthropic_responses_api.py b/tests/llm_responses_api_testing/test_anthropic_responses_api.py
index 3eabc0c15bc..1e24457145f 100644
--- a/tests/llm_responses_api_testing/test_anthropic_responses_api.py
+++ b/tests/llm_responses_api_testing/test_anthropic_responses_api.py
@@ -4,6 +4,10 @@ import pytest
 import asyncio
 from typing import Optional
 from unittest.mock import patch, AsyncMock
+from litellm.responses.litellm_completion_transformation.handler import LiteLLMCompletionTransformationHandler
+from litellm.responses.litellm_completion_transformation.transformation import LiteLLMCompletionResponsesConfig
+from litellm.types.utils import ModelResponse
+
 
 sys.path.insert(0, os.path.abspath("../.."))
 import litellm
@@ -103,4 +107,33 @@ def test_multiturn_tool_calls():
     print("follow_up_response=", follow_up_response)
         
 
-    
\ No newline at end of file
+
+
+@pytest.mark.asyncio
+async def test_async_response_api_handler_merges_trace_id_without_error():
+    handler = LiteLLMCompletionTransformationHandler()
+
+    async def fake_session_handler(previous_response_id, litellm_completion_request):
+        litellm_completion_request["litellm_trace_id"] = "session-trace"
+        return litellm_completion_request
+
+    with patch.object(
+        LiteLLMCompletionResponsesConfig,
+        "async_responses_api_session_handler",
+        side_effect=fake_session_handler,
+    ):
+        with patch("litellm.acompletion", new_callable=AsyncMock) as mock_acompletion:
+            mock_acompletion.return_value = ModelResponse(
+                id="id", created=0, model="test", object="chat.completion", choices=[]
+            )
+            await handler.async_response_api_handler(
+                litellm_completion_request={"model": "test"},
+                request_input="hi",
+                responses_api_request={"previous_response_id": "123"},
+                litellm_trace_id="original-trace",
+            )
+            # ensure acompletion called once with merged trace_id
+            assert mock_acompletion.call_count == 1
+            assert (
+                mock_acompletion.call_args.kwargs["litellm_trace_id"] == "session-trace"
+            )
\ No newline at end of file
diff --git a/tests/openai_endpoints_tests/test_e2e_openai_responses_api.py b/tests/openai_endpoints_tests/test_e2e_openai_responses_api.py
index c2e17d3ef46..e637066d2f9 100644
--- a/tests/openai_endpoints_tests/test_e2e_openai_responses_api.py
+++ b/tests/openai_endpoints_tests/test_e2e_openai_responses_api.py
@@ -119,3 +119,12 @@ def test_bad_request_bad_param_error():
         client.responses.create(
             model="gpt-4o", input="This should fail", temperature=2000
         )
+
+def test_anthropic_with_responses_api():
+    client = get_test_client()
+    response = client.responses.create(
+        model="anthropic/claude-3-5-sonnet-20240620", 
+        input="just respond with the word 'ping'",
+        previous_response_id="hi",
+    )
+    print("anthropic response=", response)