ChatOllama

Instantiate:

from langchain_ollama import ChatOllama

model = ChatOllama(
    model="gpt-oss:20b",
    validate_model_on_init=True,
    temperature=0.8,
    num_predict=256,
    # other params ...
)

Invoke:

messages = [
    ("system", "You are a helpful translator. Translate the user sentence to French."),
    ("human", "I love programming."),
]
model.invoke(messages)

AIMessage(content='J'adore le programmation. (Note: "programming" can also refer to the act of writing code, so if you meant that, I could translate it as "J'adore programmer". But since you didn\'t specify, I assumed you were talking about the activity itself, which is what "le programmation" usually refers to.)', response_metadata={'model': 'llama3', 'created_at': '2024-07-04T03:37:50.182604Z', 'message': {'role': 'assistant', 'content': ''}, 'done_reason': 'stop', 'done': True, 'total_duration': 3576619666, 'load_duration': 788524916, 'prompt_eval_count': 32, 'prompt_eval_duration': 128125000, 'eval_count': 71, 'eval_duration': 2656556000}, id='run-ba48f958-6402-41a5-b461-5e250a4ebd36-0')

Stream:

for chunk in model.stream("Return the words Hello World!"):
    print(chunk.text, end="")

content='Hello' id='run-327ff5ad-45c8-49fe-965c-0a93982e9be1'
content=' World' id='run-327ff5ad-45c8-49fe-965c-0a93982e9be1'
content='!' id='run-327ff5ad-45c8-49fe-965c-0a93982e9be1'
content='' response_metadata={'model': 'llama3', 'created_at': '2024-07-04T03:39:42.274449Z', 'message': {'role': 'assistant', 'content': ''}, 'done_reason': 'stop', 'done': True, 'total_duration': 411875125, 'load_duration': 1898166, 'prompt_eval_count': 14, 'prompt_eval_duration': 297320000, 'eval_count': 4, 'eval_duration': 111099000} id='run-327ff5ad-45c8-49fe-965c-0a93982e9be1'

stream = model.stream(messages)
full = next(stream)
for chunk in stream:
    full += chunk
full

AIMessageChunk(
    content='Je adore le programmation.(Note: "programmation" is the formal way to say "programming" in French, but informally, people might use the phrase "le développement logiciel" or simply "le code")',
    response_metadata={
        "model": "llama3",
        "created_at": "2024-07-04T03:38:54.933154Z",
        "message": {"role": "assistant", "content": ""},
        "done_reason": "stop",
        "done": True,
        "total_duration": 1977300042,
        "load_duration": 1345709,
        "prompt_eval_duration": 159343000,
        "eval_count": 47,
        "eval_duration": 1815123000,
    },
    id="run-3c81a3ed-3e79-4dd3-a796-04064d804890",
)

Async:

await model.ainvoke("Hello how are you!")

AIMessage(
    content="Hi there! I'm just an AI, so I don't have feelings or emotions like humans do. But I'm functioning properly and ready to help with any questions or tasks you may have! How can I assist you today?",
    response_metadata={
        "model": "llama3",
        "created_at": "2024-07-04T03:52:08.165478Z",
        "message": {"role": "assistant", "content": ""},
        "done_reason": "stop",
        "done": True,
        "total_duration": 2138492875,
        "load_duration": 1364000,
        "prompt_eval_count": 10,
        "prompt_eval_duration": 297081000,
        "eval_count": 47,
        "eval_duration": 1838524000,
    },
    id="run-29c510ae-49a4-4cdd-8f23-b972bfab1c49-0",
)

async for chunk in model.astream("Say hello world!"):
    print(chunk.content)

HEL
LO
WORLD
!

messages = [("human", "Say hello world!"), ("human", "Say goodbye world!")]
await model.abatch(messages)

[
    AIMessage(
        content="HELLO, WORLD!",
        response_metadata={
            "model": "llama3",
            "created_at": "2024-07-04T03:55:07.315396Z",
            "message": {"role": "assistant", "content": ""},
            "done_reason": "stop",
            "done": True,
            "total_duration": 1696745458,
            "load_duration": 1505000,
            "prompt_eval_count": 8,
            "prompt_eval_duration": 111627000,
            "eval_count": 6,
            "eval_duration": 185181000,
        },
        id="run-da6c7562-e25a-4a44-987a-2c83cd8c2686-0",
    ),
    AIMessage(
        content="It's been a blast chatting with you! Say goodbye to the world for me, and don't forget to come back and visit us again soon!",
        response_metadata={
            "model": "llama3",
            "created_at": "2024-07-04T03:55:07.018076Z",
            "message": {"role": "assistant", "content": ""},
            "done_reason": "stop",
            "done": True,
            "total_duration": 1399391083,
            "load_duration": 1187417,
            "prompt_eval_count": 20,
            "prompt_eval_duration": 230349000,
            "eval_count": 31,
            "eval_duration": 1166047000,
        },
        id="run-96cad530-6f3e-4cf9-86b4-e0f8abba4cdb-0",
    ),
]

JSON mode:

json_model = ChatOllama(format="json")
json_model.invoke(
    "Return a query for the weather in a random location and time of day with two keys: location and time_of_day. "
    "Respond using JSON only."
).content

'{"location": "Pune, India", "time_of_day": "morning"}'

Tool Calling:

from langchain_ollama import ChatOllama
from pydantic import BaseModel, Field

class Multiply(BaseModel):
    a: int = Field(..., description="First integer")
    b: int = Field(..., description="Second integer")

ans = await chat.invoke("What is 45*67")
ans.tool_calls

[
    {
        "name": "Multiply",
        "args": {"a": 45, "b": 67},
        "id": "420c3f3b-df10-4188-945f-eb3abdb40622",
        "type": "tool_call",
    }
]

Thinking / Reasoning: You can enable reasoning mode for models that support it by setting the reasoning parameter to True in either the constructor or the invoke/stream methods. This will enable the model to think through the problem and return the reasoning process separately in the additional_kwargs of the response message, under reasoning_content.

If reasoning is set to None, the model will use its default reasoning behavior, and any reasoning content will not be captured under the reasoning_content key, but will be present within the main response content as think tags (<think> and </think>).

from langchain_ollama import ChatOllama

model = ChatOllama(
    model="deepseek-r1:8b",
    validate_model_on_init=True,
    reasoning=True,
)

model.invoke("how many r in the word strawberry?")

# or, on an invocation basis:

model.invoke("how many r in the word strawberry?", reasoning=True)
# or model.stream("how many r in the word strawberry?", reasoning=True)

# If not provided, the invocation will default to the ChatOllama reasoning
# param provided (None by default).

AIMessage(content='The word "strawberry" contains **three \'r\' letters**. Here\'s a breakdown for clarity:\n\n- The spelling of "strawberry" has two parts ... be 3.\n\nTo be thorough, let\'s confirm with an online source or common knowledge.\n\nI can recall that "strawberry" has: s-t-r-a-w-b-e-r-r-y — yes, three r\'s.\n\nPerhaps it\'s misspelled by some, but standard is correct.\n\nSo I think the response should be 3.\n'}, response_metadata={'model': 'deepseek-r1:8b', 'created_at': '2025-07-08T19:33:55.891269Z', 'done': True, 'done_reason': 'stop', 'total_duration': 98232561292, 'load_duration': 28036792, 'prompt_eval_count': 10, 'prompt_eval_duration': 40171834, 'eval_count': 3615, 'eval_duration': 98163832416, 'model_name': 'deepseek-r1:8b'}, id='run--18f8269f-6a35-4a7c-826d-b89d52c753b3-0', usage_metadata={'input_tokens': 10, 'output_tokens': 3615, 'total_tokens': 3625})