Deepseek-v3.2 throws an error when running react_agent with extra_body={"thinking": {"type": "enabled"}}. #6521
Description
Checked other resources
- This is a bug, not a usage question. For questions, please use the LangChain Forum (https://forum.langchain.com/).
- I added a clear and detailed title that summarizes the issue.
- I read what a minimal reproducible example is (https://stackoverflow.com/help/minimal-reproducible-example).
- I included a self-contained, minimal example that demonstrates the issue INCLUDING all the relevant imports. The code run AS IS to reproduce the issue.
Example Code
def make_llm():
api_key = "xxx"
api_base = "xxx"
model = "deepseek/deepseek-v3.2"
return ChatOpenAI(base_url=api_base, api_key=api_key, model=model, extra_body={"thinking": {"type": "enabled"}})Error Message and Stack Trace (if applicable)
2025-12-01 21:51:30 - INFO - logger_v2 - aid=- wsid=- - [Step 2] ────────────────────────────────────────
2025-12-01 21:51:30 - INFO - logger_v2 - aid=- wsid=- - [RunLoop] ❌ Error: thread_id=test_session_100142336476783, error=Error code: 400 - {'message': 'unknown error in the model inference server trace_id: 86c6abebf34bcbaae44608dac30d9121', 'type': 'api_error'}
Traceback (most recent call last):
File "e:\multi-turn-agent\agent\react_agent.py", line 599, in _execute_with_langsmith_tracking
return await self._process_agent_stream(initial, step_count)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "e:\multi-turn-agent\agent\react_agent.py", line 634, in _process_agent_stream
async for event in self._agent.astream(initial, config=self.get_fresh_config()):
File "E:\anaconda\envs\multi_turn_agent\Lib\site-packages\langgraph\pregel\main.py", line 2939, in astream
async for _ in runner.atick(
File "E:\anaconda\envs\multi_turn_agent\Lib\site-packages\langgraph\pregel\_runner.py", line 295, in atick
await arun_with_retry(
File "E:\anaconda\envs\multi_turn_agent\Lib\site-packages\langgraph\pregel\_retry.py", line 137, in arun_with_retry
return await task.proc.ainvoke(task.input, config)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "E:\anaconda\envs\multi_turn_agent\Lib\site-packages\langgraph\_internal\_runnable.py", line 706, in ainvoke
input = await asyncio.create_task(
^^^^^^^^^^^^^^^^^^^^^^^^^^
File "E:\anaconda\envs\multi_turn_agent\Lib\site-packages\langgraph\_internal\_runnable.py", line 465, in ainvoke
ret = await asyncio.create_task(coro, context=context)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "E:\anaconda\envs\multi_turn_agent\Lib\site-packages\langgraph\prebuilt\chat_agent_executor.py", line 655, in acall_model
response = cast(AIMessage, await static_model.ainvoke(model_input, config)) # type: ignore[union-attr]
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "E:\anaconda\envs\multi_turn_agent\Lib\site-packages\langchain_core\runnables\base.py", line 3290, in ainvoke
input_ = await coro_with_context(part(), context, create_task=True)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "E:\anaconda\envs\multi_turn_agent\Lib\site-packages\langchain_core\runnables\base.py", line 5723, in ainvoke
return await self.bound.ainvoke(
^^^^^^^^^^^^^^^^^^^^^^^^^
File "E:\anaconda\envs\multi_turn_agent\Lib\site-packages\langchain_core\language_models\chat_models.py", line 417, in ainvoke
llm_result = await self.agenerate_prompt(
^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "E:\anaconda\envs\multi_turn_agent\Lib\site-packages\langchain_core\language_models\chat_models.py", line 1034, in agenerate_prompt
return await self.agenerate(
^^^^^^^^^^^^^^^^^^^^^
File "E:\anaconda\envs\multi_turn_agent\Lib\site-packages\langchain_core\language_models\chat_models.py", line 992, in agenerate
raise exceptions[0]
File "E:\anaconda\envs\multi_turn_agent\Lib\site-packages\langchain_core\language_models\chat_models.py", line 1162, in _agenerate_with_cache
result = await self._agenerate(
^^^^^^^^^^^^^^^^^^^^^^
File "E:\anaconda\envs\multi_turn_agent\Lib\site-packages\langchain_openai\chat_models\base.py", line 824, in _agenerate
response = await self.async_client.create(**payload)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "E:\anaconda\envs\multi_turn_agent\Lib\site-packages\openai\resources\chat\completions\completions.py", line 2028, in create
return await self._post(
^^^^^^^^^^^^^^^^^
File "E:\anaconda\envs\multi_turn_agent\Lib\site-packages\openai\_base_client.py", line 1748, in post
return await self.request(cast_to, opts, stream=stream, stream_cls=stream_cls)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "E:\anaconda\envs\multi_turn_agent\Lib\site-packages\openai\_base_client.py", line 1555, in request
raise self._make_status_error_from_response(err.response) from None
openai.BadRequestError: Error code: 400 - {'message': 'unknown error in the model inference server trace_id: 86c6abebf34bcbaae44608dac30d9121', 'type': 'api_error'}
During task with name 'agent' and id 'b56f4fa6-3273-8be6-3c92-f7e01e1707c1'Description
Deepseek-v3.2 throws an error when running react_agent with extra_body={"thinking": {"type": "enabled"}}.
langchain==0.3.27
langchain-community==0.3.27
langchain-core==0.3.74
langchain-openai==0.2.9
langchain-text-splitters==0.3.9
langgraph==0.6.4
langgraph-checkpoint==2.1.1
langgraph-prebuilt==0.6.4
langgraph-sdk==0.2.0
langsmith==0.4.13
System Info
python react_agent.py