The maximum number of tokens that can be generated in the chat completion. The total length of input tokens and generated tokens is limited by the model's context length. This value is now deprecated in favor of 'max_completion_tokens' parameter.
max_tokens: int | None = None