Build a live translation broadcast app with the Gemini Live API and LiveKit
Gemini 3.5 live translation model enables real-time multi-language broadcast via API
“This is using the Gemini API together with Life Kit and Google Cloud Run to deploy an application that can broadcast speech in many, many different languages using Gemini live translate.”
Google demoed a live translation broadcast app built on the new Gemini 3.5 live translation model, combining the Gemini API with LiveKit and Google Cloud Run for real-time multi-language audio streaming. The architecture uses WebRTC for audio and data channels for captions, with session management that shares translation sessions across listeners of the same language. The demo is open source on GitHub, but this is a developer tutorial rather than a strategic announcement.