kapynAI / Models

Gemini Live audio

Gemini 3.8 Live is a new speech-to-speech model featuring real-time interruption capabilities. Google released Gemini 3.8 Live and Gemini 3.8 Live Extended Thinking to compete directly with OpenAI's voice conversation models. Developers can test these low-latency audio capabilities through browser-based voice interfaces with customizable system prompts and voice presets.

Simon Willison·Sep 15, 2026

Opening Kapyn…