Gemini 3.5 Live Translate

Gemini 3.5 Live Translate, announced on 9 June 2026, is Google's dedicated speech-to-speech translation model and the closest thing the industry has yet produced to a machine simultaneous interpreter. It detects the spoken language by itself and begins speaking the translation while the original sentence is still under way, trailing the speaker by a few seconds rather than waiting for a pause. What separates it from a transcription-plus-synthesis pipeline is that it carries the delivery across: pacing, intonation and pitch survive the crossing, so the translated voice keeps some of the shape of the original. Google put it in three very different places at once — the Live API and AI Studio for developers, a private preview inside Google Meet for selected Workspace customers, and the Google Translate app on Android and iOS for everyone else — which says plainly that this is aimed at meetings and travel, not at research. Priced by the minute-equivalent at $3.50 per million audio input tokens and $21 per million output, it is dearer than Google's conversational Live model but still well below OpenAI's flagship voice pricing.

▸Videos

No videos yet.

▸Reviews