Gemini 3.5 Live Translate

A Google audio model that translates speech to speech in near real time across 70+ languages, rolling out in the Google Translate app.

Translates speech continuously as you talk instead of waiting for each speaker to finish, keeping the conversation roughly in sync.
Preserves the speaker's intonation, pacing, and pitch in the translated voice and auto-detects the language being spoken.
Rolling out globally in the Google Translate app on Android and iOS, with an Android listening mode that plays translations through the phone's earpiece.
A private preview opens for select business Google Workspace customers this month, with broader rollout later in the year.
Developers can build voice translation apps through the Gemini Live API via partners like LiveKit and Pipecat.