thinking_level

Supported values:

• 'low': Minimizes latency and cost.
• 'medium': Balances latency/cost with reasoning depth.
• 'high': Maximizes reasoning depth.

thinking_budget is deprecated for Gemini 3+ models. If both parameters are provided, thinking_level takes precedence.

If left unspecified, the model's default thinking level is used. For Gemini 3+, this defaults to 'high'.