ChatConfig

atoti_ai_openai.ChatConfig(
    *,
    model: str,
    temperature: float | None = None,
    top_p: float | None = None,
    max_tokens: int | None = None,
    max_completion_tokens: int | None = None,
    seed: int | None = None,
    tool_choice: str | None = None,
    parallel_tool_calls: bool | None = None,
    extra_body: Mapping[str, bool | int | float | str] | None = None,
)

Chat config for OpenAI. See the corresponding Javadoc for more details.

​

Attributes

​

model

The ID of the model to use, for example "gpt-4o".

​

temperature

Controls randomness in responses from 0.0 (deterministic) to 2.0 (creative). Not supported by reasoning models (e.g. o1, o3, gpt-5).

​

top_p

Nucleus sampling parameter. Controls diversity via cumulative probability. Not supported by reasoning models (e.g. o1, o3, gpt-5).

​

max_tokens

Maximum number of tokens to generate in the response. For non-reasoning models. Mutually exclusive with max_completion_tokens.

​

max_completion_tokens

Maximum number of tokens to generate for reasoning models (e.g. o1, o3). Mutually exclusive with max_tokens.

​

seed

Seed for deterministic sampling (Beta).

​

tool_choice

Tool/function calling behavior ("none", "auto", or a specific function name).

​

parallel_tool_calls

Enable parallel function calling during tool use.

​

extra_body

Additional parameters for OpenAI-compatible servers.