openai/openai-python · error · ValueError

`{tool['function']['name']}` is not strict. Only `strict` fu

Error message

`{tool['function']['name']}` is not strict. Only `strict` function tools can be auto-parsed

What it means

Auto-parsing of tool-call arguments requires strict JSON-schema function tools (function.strict == True) so the arguments are guaranteed valid JSON. The validator rejects any function tool whose 'strict' field is not exactly true.

Source

Thrown at src/openai/lib/_parsing/_completions.py:79

    return [t for t in tools if is_strict_chat_completion_tool_param(t)]


def validate_input_tools(
    tools: Iterable[ChatCompletionToolUnionParam] | Omit = omit,
) -> Iterable[ChatCompletionFunctionToolParam] | Omit:
    if not is_given(tools):
        return omit

    for tool in tools:
        if tool["type"] != "function":
            raise ValueError(
                f"Currently only `function` tool types support auto-parsing; Received `{tool['type']}`",
            )

        strict = tool["function"].get("strict")
        if strict is not True:
            raise ValueError(
                f"`{tool['function']['name']}` is not strict. Only `strict` function tools can be auto-parsed"
            )

    return cast(Iterable[ChatCompletionFunctionToolParam], tools)


def parse_chat_completion(
    *,
    response_format: type[ResponseFormatT] | completion_create_params.ResponseFormat | Omit,
    input_tools: Iterable[ChatCompletionToolUnionParam] | Omit,
    chat_completion: ChatCompletion | ParsedChatCompletion[object],
) -> ParsedChatCompletion[ResponseFormatT]:
    if is_given(input_tools):
        input_tools = [t for t in input_tools]
    else:
        input_tools = []

    choices: list[ParsedChoice[ResponseFormatT]] = []

View on GitHub (pinned to 9917c6e28e)

Solutions

1. Set "strict": True in the function tool definition
2. Ensure the tool's parameters JSON schema is strict-compatible (additionalProperties: false, all fields required) — to_strict_json_schema can do this from a Pydantic model
3. Use pydantic_function_tool(MyModel) to generate a compliant strict tool param

Example fix

# before
tools = [{"type":"function","function":{"name":"get_weather","parameters":{...}}}]
# after
tools = [{"type":"function","function":{"name":"get_weather","strict":True,"parameters":strict_schema}}]

for t in tools:
    if t["type"] == "function" and t["function"].get("strict") is not True:
        raise ValueError(f"tool {t['function']['name']} must be strict")

def is_strict_function_tool(t: dict) -> bool:
    return t.get("type") == "function" and t["function"].get("strict") is True

When it happens

Trigger: Passing a function tool without strict: true (or strict: false) to client.chat.completions.parse.

Common situations: Copying tool definitions from completions.create where strict is optional; manually building tool dicts and omitting 'strict'; strict set to truthy non-True value.

Related errors

• Currently only `function` tool types support auto-parsing; R
• Unable to automatically parse response format type {response
• Expected Content-Type response header to be `application/jso
• invalid datetime format
• invalid date format

AI-assisted analysis of openai/openai-python@9917c6e28e (2026-08-28). Data as JSON: /api/errors/c6a686af62b6fb21. Report an issue: GitHub.