Gemini 2.5 introduces advanced audio capabilities, enabling real-time audio dialog with natural conversation, style control, and tool integration. It also features controllable text-to-speech, allowing users to dictate style, tone, and expression in generated audio. These advancements empower developers to create richer applications through the Gemini API. The technology also has multilingual support and safety measures.
https://deepmind.google/discover/blog/advanced-audio-dialog-and-generation-with-gemini-25/