Support for adding functions dynamically during a stream just like openai

Hi do we have the support to update functions of the AI agent, while we are having voice stream with the gemini?