Gemini 3.5 Live Translate

2026-06-09 · Publisher : Google · Domain : Artificial Intelligence

Google launches Gemini 3.5 Live Translate, a near-real-time speech-to-speech voice translation model, with automatic detection of more than 70 languages and preservation of the speaker's intonation.

What changes

  • Continuous generation of the translated speech, without waiting for the end of the sentence, a few seconds behind the speaker.
  • Automatic detection of more than 70 languages, robustness to noise.
  • Preserves the intonation, rhythm and pitch of the voice.
  • Google Meet: going from 5 to more than 70 languages and from translation into English only to more than 2000 combinations (private preview).

### Specifications - Availability: Live API and AI Studio in public preview; Google Meet in private preview; Google Translate app (Android, iOS). - Exact API identifier: not verified. - Audio output marked with SynthID. ### Pricing Current pricing page: input 3.50 USD per million tokens or 0.0053 USD per minute; output 21.00 USD per million tokens or 0.0315 USD per minute. Pricing on the day of the announcement not verified. ### Benchmarks No score published in the blog post. ### Key takeaways A simultaneous translation building block for multilingual calls and meetings, still in preview for developers.

Links

Tags : Audio, Google, Gemini, Traduction, Voix