Google has expanded real-time speech translation in Google Meet to 63 languages, the company said in a Workspace Updates post. The Gemini-powered feature now works across web, Android, iOS and Google Meet hardware devices, letting participants speak different languages in the same call and hear translated audio almost instantly.
Key points
- Google Meet's Gemini-powered speech translation now covers 63 languages across web, Android, iOS and Meet hardware.
- Participants can mix any combination of supported languages in one meeting, with automatic language detection replacing the old single-pair system.
- Usage caps range from 90 minutes per meeting on base Workspace and AI Pro plans to 8 hours on AI Expanded and Ultra plans.
- The feature does not work during live streams, recordings, encrypted meetings, dial-in calls or interoperability calls.
- Meetings require an eligible Google Workspace or Google AI plan holder as organizer, or as a participant for one-on-one calls.
The update replaces a far more limited system that required every participant to agree on one language pair for an entire meeting. For anyone running international teams or mixed-language calls, that single change removes a major friction point: a French-speaking colleague joining an English-Spanish meeting previously got no translation at all.
How The New Translation System Works
Speech translation now detects which language a person is speaking automatically, even if they switch mid-sentence or multiple speakers use different languages in the same meeting. Each listener hears audio translated into whichever of the 63 supported languages they prefer, and synchronized captions display the translated text alongside the audio. Translations lag the original speech by a few seconds to maintain accuracy.
Participants turn translation on themselves through the meeting toolbar rather than waiting for a host to set it up, according to the Google Workspace Updates announcement. Anyone who already understands certain languages can opt out of translation for those languages specifically, and users can change their preferences at any point during a call in Meet settings. Google says it does not store voice data or use it to train models.
Who Can Use It And For How Long
Access still depends on subscription tier. Scheduled meetings need an organizer on an eligible Google Workspace or Google AI plan, which then extends translation to every participant, including outside guests who have no subscription of their own. For one-on-one calls, only one of the two participants needs an eligible plan.
Time limits vary by plan:
- Google AI Pro, Workspace Business Standard and Plus, Enterprise Standard and Plus, Frontline Plus, and AI Pro for Education: 90 minutes per meeting
- Google AI Ultra and AI Expanded Access: 8 hours per meeting
Free personal Google accounts cannot use speech translation or access translated captions in meetings they organize, which keeps the feature largely in paid and business tiers for now.
Where The Feature Still Falls Short
Several common scenarios are excluded entirely. Speech translation does not work during live streams, recorded meetings, client-side encrypted meetings, dial-in calls, or calls using third-party interoperability tools. It is also unavailable when a meeting uses adaptive audio, and access is further limited by region and age restrictions that Google has not detailed beyond confirming they exist.
Conference rooms running Google Meet hardware can listen to translated speech coming from remote participants, but audio originating in the room itself is not translated outward to others. That one-way limitation means hybrid meetings with in-room and remote attendees still need some manual coordination if everyone needs to be understood.
Part Of A Broader Push Into AI Meeting Tools
Speech translation first became generally available for businesses in January 2026, initially covering just English and five European languages through a single fixed pair per call. The 63-language expansion, powered by what outlets have described as Gemini 3.5 Live Translate, marks the biggest jump in coverage since launch.
The update adds to a run of AI features Google has layered into its conferencing app, including the note-taking tools covered in DroidFeats’ report on Meet gaining automatic transcription. Google has also pushed Meet into new places entirely, such as the audio-only access detailed in DroidFeats’ coverage of Google Meet arriving on Apple CarPlay.
Google has not published a full list of the 63 supported languages or specified when the rollout will reach every eligible account. Based on how the company has handled similar staged releases, availability could take up to two weeks to reach all users.
