Google unveiled a new live translation feature for its Google Meet video conferencing service at its I/O event. The technology, powered by Gemini AI, converts a user’s speech into the language spoken by the other person, while preserving intonation and emotion. The feature is currently available in beta mode.
Image source: Google
In the demo video, one participant spoke English, the other Spanish. After activating Gemini’s translation feature, each participant heard their partner speak in their own language with incredible intonation accuracy, The Verge writes.
The system currently only supports English and Spanish. However, Google has already announced that Italian, German and Portuguese will soon be added. According to the company, the new languages will appear in the coming weeks.
A similar feature was launched by Microsoft earlier this year in its Teams collaboration app, but Google is focusing on more natural-sounding translation, emphasizing that the AI doesn’t just translate text, but adapts the user’s voice so that their speech sounds organic.
The feature is being tested in beta mode for Google AI Pro subscribers and the new $249 per month AI Ultra plan. The final version may appear later.