Google today announced Gemini 3.5 Live Translate as its latest model for live speech-to-speech translation.

This model can detect over 70 languages and generate “smooth, natural-sounding translated speech” that preserves intonation, pacing, and pitch.

Compared to turn-by-turn translation systems that “wait for the speaker to finish speaking before responding,” this model continuously generates speech. This results in “fluid audio without awkward pauses and stays just a few seconds behind the speaker throughout the session.”

…3.5 Live Translate generates speech continuously, balancing the trade-off between waiting for context to improve quality and translating immediately to stay in sync with the speaker.

Gemini 3.5 Live Translate is rolling out now to Google Translate on Android and iOS. Tap “Live translate” in the bottom-left corner while wearing a pair of headphones.

The Android app is also rolling out a new “listening mode” that lets you hear translations through the phone’s earpiece when you don’t have headphones immediately available. You can “simply hold your phone to your ear just like a regular call.”

Gemini 3.5 Live Translate is also coming to Google Meet. Speech translation in the video calling app previously supported just five languages. The model’s support for 70+ languages unlocks “2000+ language combinations in one meeting, expanding from the previous state of only translating to and from English.” On the web, there’s a new button in the row of controls to immediately start speech translation.

This month, it’s rolling out in “private preview for select business Google Workspace customers.” A broader rollout is coming later this year.

Gemini 3.5 Live Translate is also available in public preview for developers via the Gemini Live API and Google AI Studio.

On the safety front, all generated audio is watermarked with SynthID:

This imperceptible watermark is woven directly into the audio output, ensuring AI-generated content remains detectable to help prevent misinformation