Polyman Documentation - v3.1.0
    Preparing search index...

    Interface ChatRequest

    A request for one model turn.

    interface ChatRequest {
        model: string;
        system: string;
        messages: Message[];
        tools: ToolSpec[];
        maxOutputTokens?: number;
        temperature?: number;
        reasoningEffort?: "low" | "medium" | "high";
    }
    Index
    model: string

    Model id, as the provider names it.

    system: string

    System prompt.

    messages: Message[]

    The conversation so far.

    tools: ToolSpec[]

    Tools the model may call. Empty when none are offered.

    maxOutputTokens?: number

    Upper bound on generated tokens.

    temperature?: number

    Sampling temperature.

    reasoningEffort?: "low" | "medium" | "high"

    How much the model should think, for models that support it.