microsoft/autogen · error · ValueError

model_info is required when model name is not a valid OpenAI

Error message

model_info is required when model name is not a valid OpenAI model

What it means

When neither model_capabilities nor model_info is passed, OllamaChatCompletionClient tries to look up the model name in the autogen_ext.models._model_info registry (the OpenAI model table). A KeyError from that lookup is re-raised as this ValueError: custom/Ollama model names are not in the OpenAI registry, so the caller must supply model_info explicitly.

Source

Thrown at python/packages/autogen-ext/src/autogen_ext/models/ollama/_ollama_client.py:473

    create_args: Dict[str, Any]


class BaseOllamaChatCompletionClient(ChatCompletionClient):
    def __init__(
        self,
        client: AsyncClient,
        *,
        create_args: Dict[str, Any],
        model_capabilities: Optional[ModelCapabilities] = None,  # type: ignore
        model_info: Optional[ModelInfo] = None,
    ):
        self._client = client
        self._model_name = create_args["model"]
        if model_capabilities is None and model_info is None:
            try:
                self._model_info = _model_info.get_info(create_args["model"])
            except KeyError as err:
                raise ValueError("model_info is required when model name is not a valid OpenAI model") from err
        elif model_capabilities is not None and model_info is not None:
            raise ValueError("model_capabilities and model_info are mutually exclusive")
        elif model_capabilities is not None and model_info is None:
            warnings.warn("model_capabilities is deprecated, use model_info instead", DeprecationWarning, stacklevel=2)
            info = cast(ModelInfo, model_capabilities)
            info["family"] = ModelFamily.UNKNOWN
            self._model_info = info
        elif model_capabilities is None and model_info is not None:
            self._model_info = model_info

        self._resolved_model: Optional[str] = None
        self._model_class: Optional[str] = None
        if "model" in create_args:
            self._resolved_model = create_args["model"]
            self._model_class = _model_info.resolve_model_class(create_args["model"])

        if (
            not self._model_info["json_output"]

View on GitHub (pinned to 027ecf0a37)

Solutions

Pass model_info describing the model: OllamaChatCompletionClient(model='llama3.2', model_info={'vision': True, 'function_calling': True, 'json_output': True, 'family': ModelFamily.UNKNOWN, 'structured_output': True})
Or use a name that exists in the registry (an OpenAI-known model name)
Build the dict from ModelInfo fields; include 'family' to satisfy the type

Example fix

# before
client = OllamaChatCompletionClient(model="llama3.2:latest")

# after
client = OllamaChatCompletionClient(
    model="llama3.2:latest",
    model_info={"vision": True, "function_calling": True, "json_output": True, "structured_output": True, "family": ModelFamily.UNKNOWN},
)

Defensive patterns

Strategy: validation

Validate before calling

from autogen_ext.models._model_info import ModelInfo

def with_model_info(cfg: dict) -> dict:
    try:
        ModelInfo.get_info(cfg["model"])
    except KeyError:
        cfg.setdefault("model_info", {
            "vision": False, "function_calling": True,
            "json_output": True, "structured_output": True,
            "family": ModelFamily.UNKNOWN,
        })
    return cfg

client = OllamaChatCompletionClient(**with_model_info(cfg))

Try / catch

try:
    client = OllamaChatCompletionClient(model=name)
except ValueError as e:
    if "model_info is required" in str(e):
        client = OllamaChatCompletionClient(model=name, model_info=default_info)
    else:
        raise

Prevention

Always pass an explicit model_info when using local Ollama tags
Keep a small dict of model_info per model tag in your config
On autogen-ext upgrades, re-check which Ollama tags the bundled registry knows

When it happens

Trigger: OllamaChatCompletionClient(model='llama3.2:latest') with no model_info; model='mistral-nemo' or any name not present in the OpenAI model info table; version drift where a newly shipped Ollama tag is missing from the bundled registry.

Common situations: First-time use with locally pulled Ollama tags; upgrading Ollama to a new model release before autogen-ext's model table updates; copying example code that omitted model_info.

Related errors

AI-assisted analysis of microsoft/autogen@027ecf0a37 (2026-08-15). Data as JSON: /api/errors/b11618c348482b12. Report an issue: GitHub.