[LangGraph + Ollama] Agent using local model (qwen2.5) returns AIMessage(content='') even when tool responds correctly

[matiasdev30]

[LangGraph + Ollama] Agent using local model (qwen2.5) returns AIMessage(content='') even when tool responds correctly

I’m using create_react_agent from langgraph.prebuilt with a local model served via Ollama (qwen2.5), and the agent consistently returns an AIMessage with an empty content field — even though the tool returns a valid string.

Code

from langgraph.prebuilt import create_react_agent
from langchain_ollama import ChatOllama

model = ChatOllama(model="qwen2.5")

def search(query: str):
    """Call to surf the web."""
    if "sf" in query.lower() or "san francisco" in query.lower():
        return "It's 60 degrees and foggy."
    return "It's 90 degrees and sunny."

agent = create_react_agent(model=model, tools=[search])

response = agent.invoke(
    {},
    {"messages": [{"role": "user", "content": "what is the weather in sf"}]}
)
print(response)

Output

{
  'messages': [
    AIMessage(
      content='', 
      additional_kwargs={}, 
      response_metadata={
        'model': 'qwen2.5',
        'created_at': '2025-04-24T09:13:29.983043Z',
        'done': True,
        'done_reason': 'load',
        'total_duration': None,
        'load_duration': None,
        'prompt_eval_count': None,
        'prompt_eval_duration': None,
        'eval_count': None,
        'eval_duration': None,
        'model_name': 'qwen2.5'
      },
      id='run-6a897b3a-1971-437b-8a98-95f06bef3f56-0'
    )
  ]
}

As shown above, the agent responds with an empty string, even though the search() tool clearly returns "It's 60 degrees and foggy.".

Has anyone seen this behavior? Could it be an issue with qwen2.5, langgraph.prebuilt, the Ollama config, or maybe a mismatch somewhere between them?

Any insight appreciated.
``

[dosubot[bot]]

Hello @matiasdev30! I'm Dosu, a bot here to help you solve bugs, answer questions, and assist with becoming a contributor. I'm ready to support you while you wait for a human maintainer. How can I assist you today?

The issue you're encountering, where the AIMessage returns an empty content field despite the tool returning a valid response, could be related to how the create_react_agent function and the qwen2.5 model are handling tool outputs. In LangChain, an AIMessage with an empty content field can occur when the message is used to represent a tool call with specific arguments and function names, and the primary information is contained within the additional_kwargs or tool_calls ^[1].

Here are a few things you might want to check or try:

1. Tool Output Handling: Ensure that the tool's output is being correctly captured and passed back to the agent. The invoke method should be handling tool outputs properly, but there might be a configuration or integration issue with the qwen2.5 model.

2. Model-Specific Behavior: There might be specific behavior or limitations with the qwen2.5 model when used with LangChain and Ollama. Although I couldn't find specific known issues with this model, it's worth checking if there are any model-specific configurations or updates required ^[2].

3. Agent Configuration: Double-check the configuration of the create_react_agent function and ensure that the tools are correctly bound to the model. The function should automatically handle this, but any discrepancies in setup might lead to unexpected behavior ^[3].

4. Debugging: Add logging or print statements within the search function and the agent's processing steps to trace where the output might be getting lost or not properly integrated into the AIMessage.

If these steps don't resolve the issue, consider testing with a different model or tool to isolate whether the problem is specific to the qwen2.5 model or the integration with Ollama.

To continue talking to Dosu, mention @dosu.

---

^{Help Dosu learn! Give it feedback: Great Response | Irrelevant Answer | Incorrect Sources | Too Verbose | Hallucination | Bug Report | Other}