minOutputTokens property

int? minOutputTokens
final

The minimum output-token budget the model needs (session_defaults.max_output_tokens_min) — e.g. a reasoning model is cut off mid-<think> below this. This is a FLOOR, not a cap: when you set createSession(maxOutputTokens:), keep it at or above this value (or leave the cap unset). Do NOT assume it is itself the cap to apply blindly — passing a floor into a smaller cap would truncate output.

Implementation

final int? minOutputTokens;