microsoft/autogen · error · ValueError

json_output and format cannot be set at the same time. Use j

Error message

json_output and format cannot be set at the same time. Use json_output instead.

What it means

Ollama's native API uses a 'format' parameter for output shaping. create() treats a 'format' key in create_args and the json_output parameter as mutually exclusive — setting both raises ValueError. The code then asserts response_format_value is None and adopts create_args['format'] verbatim, so ambiguity about which shaping wins is not allowed.

Source

Thrown at python/packages/autogen-ext/src/autogen_ext/models/ollama/_ollama_client.py:565

                response_format_value = "json"
            elif json_output is False:
                # Text mode.
                response_format_value = None
            elif isinstance(json_output, type) and issubclass(json_output, BaseModel):
                if response_format_value is not None:
                    raise ValueError(
                        "response_format and json_output cannot be set to a Pydantic model class at the same time. "
                        "Use json_output instead."
                    )
                # Beta client mode with Pydantic model class.
                response_format_value = json_output.model_json_schema()
            else:
                raise ValueError(f"json_output must be a boolean or a Pydantic model class, got {type(json_output)}")

        if "format" in create_args:
            # Handle the case where format is set from create_args.
            if json_output is not None:
                raise ValueError("json_output and format cannot be set at the same time. Use json_output instead.")
            assert response_format_value is None
            response_format_value = create_args["format"]
            # Remove format from create_args to prevent passing it twice.
            del create_args["format"]

        # TODO: allow custom handling.
        # For now we raise an error if images are present and vision is not supported
        if self.model_info["vision"] is False:
            for message in messages:
                if isinstance(message, UserMessage):
                    if isinstance(message.content, list) and any(isinstance(x, Image) for x in message.content):
                        raise ValueError("Model does not support vision and image was provided")

        if self.model_info["json_output"] is False and json_output is True:
            raise ValueError("Model does not support JSON output.")

        ollama_messages_nested = [to_ollama_type(m) for m in messages]
        ollama_messages = [item for sublist in ollama_messages_nested for item in sublist]

View on GitHub (pinned to 027ecf0a37)

Solutions

Remove 'format' from extra_create_args/create_args and control output shape solely via json_output
Or drop json_output and pass the native format — but the json_output path is the supported autogen API
When building create_args programmatically, create_args.pop('format', None) if you also pass json_output

Example fix

# before
await client.create(messages, json_output=True, extra_create_args={"format": "json"})

# after
await client.create(messages, json_output=True)

Defensive patterns

Strategy: validation

Validate before calling

if json_output is not None:
    create_args.pop("format", None)  # json_output owns output shaping
result = await client.create(messages, extra_create_args=create_args, json_output=json_output)

Try / catch

try:
    result = await client.create(messages, extra_create_args=extra, json_output=jo)
except ValueError as e:
    if "format cannot be set" in str(e):
        extra.pop("format", None)
        result = await client.create(messages, extra_create_args=extra, json_output=jo)
    else:
        raise

Prevention

Pick one output-shaping channel (json_output) and strip native 'format' from forwarded kwargs
When porting raw ollama-python code, whitelist the kwargs you forward rather than passing them all
Keep native Ollama options under options=..., not as top-level create args

When it happens

Trigger: create(messages, json_output=True, extra_create_args={'format': 'json'}); constructing the client with format='json' in its config and later passing json_output on individual calls; forwarding Ollama-native options dicts that include format.

Common situations: Client config containing native Ollama options (options={'format': 'json'} style) combined with the autogen-level json_output; migrating raw ollama-python call code into autogen extra_create_args without stripping format.

Related errors

AI-assisted analysis of microsoft/autogen@027ecf0a37 (2026-08-15). Data as JSON: /api/errors/61bddfe9497e9a8b. Report an issue: GitHub.