At Google I/O 2025, the tech giant unveiled a revolutionary advancement in communication technology: real-time speech translation in Google Meet that functions like having a human interpreter in your video calls.
Unlike traditional translation tools that rely on captions or robotic voices, Google's new feature performs direct audio-to-audio transformations. When activated, the system translates spoken words into the listener's preferred language while preserving the speaker's original voice characteristics, including tone, intonation, and emotional expressions. Users hear a faint version of the original speech with the translated version overlaid on top, creating an authentic conversational experience.
"The feature utilizes a robust language audio model developed by Google DeepMind, enabling seamless and natural conversations between individuals speaking different languages," Google stated during the I/O presentation. The technology leverages AudioLM, which is trained on audio data to maintain as much of the original audio quality as possible while transforming the language.
The feature is currently rolling out in beta to Google AI Pro ($19.99/month) and the newly launched AI Ultra ($249.99/month) subscribers. Importantly, only one participant in a call needs a subscription for the translation to work for everyone. Initially supporting English and Spanish, Google plans to add Italian, German, and Portuguese in the coming weeks.
The practical applications are extensive, from enabling English-speaking grandchildren to converse naturally with Spanish-speaking grandparents to facilitating seamless international business collaboration. Google confirmed that early testing for Workspace business customers will begin later this year, potentially transforming global corporate communication.
This development represents a significant leap beyond competitors' offerings. While Microsoft Teams and Skype have offered translation features in the past, Google's approach delivers a more fluid, natural experience that maintains the human element of conversation—a crucial advancement in breaking down global language barriers.