Instant translation in 63 languages with artificial intelligence

Google is expanding the artificial intelligence-supported speech translation feature of its video conferencing platform Google Meet to 63 languages. The new system, supported by Gemini 3.5 Live Translate, automatically detects the spoken language and allows participants to listen to conversations in different languages ​​in the language of their choice, in almost real time.

According to Google's announcement dated October 22, with the update, translation will be possible between different combinations of 63 supported languages ​​in the same meeting. While in the previous system, a single language pair had to be configured throughout the meeting, the new version will work with automatic language detection in meetings where different languages ​​are spoken.

THE LANGUAGE SPOKE WILL BE AUTOMATIC DETECTED

With the new feature, Google Meet will automatically detect the language spoken during the meeting and translate the audio into the preferred language of the participants. Translation will continue if a speaker changes language during the conversation or if more than one language is used in the same meeting.

Participants will be able to enable the feature individually for themselves using the translation control in the meeting toolbar. It will also be able to identify other languages ​​they understand and prevent conversations in these languages ​​from being translated.

The system will also offer simultaneous subtitle support with the translated audio. Thus, participants will be able to both listen to the speeches and follow them through the translated text.

TRANSLATION TIME VARIES DEPENDING ON SUBSCRIPTION PLAN

Google Meet's speech translation feature will be available as part of eligible Google Workspace and Google AI subscriptions. When an eligible subscriber hosts the meeting, all participants will be able to take advantage of the feature, including guests from different organizations. In one-on-one meetings, it will be sufficient for one of the participants to have an appropriate plan.

While translation time is limited to 90 minutes per meeting in basic eligible Workspace plans and Google AI Pro subscription, this time will be up to 8 hours in AI Expanded Access, AI Ultra Access and Google AI Ultra plans.

Feature; It will be available on Business Standard and Plus, Enterprise Standard and Plus, Frontline Plus, Google AI Pro for Education and specified Google AI access plans.

SOME MEETINGS WILL BE EXCLUDED

Google Meet's speech translation feature will not be available in live broadcasts, meeting recordings, client-side encrypted conversations, telephone participation, and third-party interoperability conversations. Calls using the adaptive voice feature will also not be supported.

Google stated that users can control whether their conversations are translated for others and change these preferences in Meet settings. The company also stated that it does not store voice data and does not use users' voices to train its models.

UPDATE IS AVAILABLE IN PHASE

According to Google's statement, distribution for accounts in the fast release channel started on October 22, and the feature's visibility was planned to be completed within 1 to 3 days. For accounts in the planned release channel, distribution is expected to start on November 2 and last up to 15 days.

To benefit from new features on mobile devices, the current version of the Google Meet application must be used. Speech translation will be off by default for users when the meeting starts and can be enabled by participants during the call.