Google has unveiled a groundbreaking real-time voice translation feature for its cloud-based video conferencing platform, Google Meet.
TechCrunch reported on Tuesday that this innovative feature leverages Google DeepMind’s advanced large language audio model to deliver translations that capture words, voice inflections, and emotional nuances. The company explains that this technology enables seamless communication between English-speaking grandchildren and their Spanish-speaking grandparents and facilitates real-time dialogue among employees of multinational corporations.
Google claims the minimal translation lag allows for fluid conversations even with multiple participants. The system overlays translated speech onto a faint background of the original voice, creating a natural and immersive conversational experience.
Currently in beta, this feature is available exclusively to AI subscribers and initially supports English and Spanish. Google plans to expand language options to include Italian, German, and Portuguese within the coming weeks.
The tech giant is also developing an enterprise version of this feature and intends to roll out early testing to Workspace customers later this year.