Chat completions

Request

model and a non-empty messages array are required.

{
  "model": "pro",
  "messages": [
    {"role": "system", "content": "Answer with short, direct sentences."},
    {"role": "user", "content": "Review this deployment plan."}
  ],
  "temperature": 0.2,
  "max_tokens": 1200
}

Use max_tokens to set the maximum output. Do not send both max_tokens and max_completion_tokens in the same request. MaxLabs may lower the effective output limit when the account has limited quota left. Your request can still succeed with a shorter answer.

Response

{
  "id": "completion-id",
  "object": "chat.completion",
  "model": "pro",
  "choices": [
    {
      "index": 0,
      "message": {"role": "assistant", "content": "..."},
      "finish_reason": "stop"
    }
  ],
  "usage": {
    "prompt_tokens": 84,
    "completion_tokens": 42,
    "total_tokens": 126
  }
}

Check finish_reason. A value of length means the response reached an output limit. Increase the limit only if your model and remaining quota allow it. Keep the public model value from the response; do not depend on any provider-specific identity.