How to set Ollama's think=False parameter in LangGraph Agent?

[J7503303]

After version 0.9, Ollama supports turning off the Think mode for large language models. You can disable it by entering /set nothink in the command line or by passing the parameter think=False.

I'm not sure how to set this parameter (think=False) when using Ollama within a LangGraph Agent. Does anyone know how to configure this? Please help. Thanks.

[HuyPham-NeurondAI]

Hi mate, I found 2 ways to handle this.

1. Set extract_reasoning param as True when creating an instance of ChatOllama

Like this:

llm = ChatOllama(
    model="qwen3:0.6b",
    extract_reasoning=True
)

This will make the response from llm.invoke() have a separate part of the real answer and a thinking block:

AIMessage(content='\n\nHello! How can I assist you today? 😊', additional_kwargs={'reasoning_content': '<think>\nOkay, the user just said "Hello". I need to respond appropriately. Let me start by acknowledging their greeting. A simple "Hello!" is good. Then, I can add a friendly message to keep the conversation going. Maybe something like "How can I assist you today?" to encourage further interaction. It\'s important to keep the tone positive and open-ended so they feel comfortable sharing more. I should make sure the response is clear and polite, avoiding any unnecessary details. Alright, putting it all together in a natural flow.\n</think>'}, response_metadata={'model': 'qwen3:0.6b', 'created_at': '2025-07-07T06:29:59.7201856Z', 'done': True, 'done_reason': 'stop', 'total_duration': 1765612600, 'load_duration': 23520900, 'prompt_eval_count': 9, 'prompt_eval_duration': 32823200, 'eval_count': 122, 'eval_duration': 1708222000, 'model_name': 'qwen3:0.6b'}, id='run--432ba6f0-e660-4935-b943-d4078985a21f-0', usage_metadata={'input_tokens': 9, 'output_tokens': 122, 'total_tokens': 131})

2. Set "think" param as False when invoke the llm

Like this:

response = llm.invoke("Hello", think=False)

This will disable think mode from the model and make the response faster and less token

AIMessage(content='Hello! What can I do for you today? 😊', additional_kwargs={}, response_metadata={'model': 'qwen3:0.6b', 'created_at': '2025-07-07T06:39:22.7074581Z', 'done': True, 'done_reason': 'stop', 'total_duration': 189633600, 'load_duration': 21198800, 'prompt_eval_count': 17, 'prompt_eval_duration': 13754100, 'eval_count': 13, 'eval_duration': 154161700, 'model_name': 'qwen3:0.6b'}, id='run--2c35fed6-fbf3-490b-9e31-1b0ddb1cdb0e-0', usage_metadata={'input_tokens': 17, 'output_tokens': 13, 'total_tokens': 30})