Google Meet now offers real-time speech translation, breaking language barriers for global teams and enhancing live multilingual communication
At Google I/O 2025, Google disclosed that it will implement real-time speech translation in Google Meet.
According to Google, the feature facilitates a genuine, fluid conversation with an individual who speaks a different language by utilizing a large language audio model from Google DeepMind.
In real-time, speech translation in Meet translates spoken words into the listener’s preferred language. The translation maintains the original voice, tone, and expression.
According to the technology behemoth, the new feature has diverse applications. For example, it may be implemented to facilitate communication between Spanish-speaking grandparents and their English-speaking descendants.
Alternatively, it could be implemented by organizations that operate in multiple regions to facilitate the communication and collaboration of global counterparts in real-time.
According to Google, the latency for speech translation is extremely low, enabling simultaneous communication of multiple individuals, a feat that has been unattainable until now.
The original audio of the individual on the other side will be faintly audible, with the translated speech superimposed on top.
Google Meet will commence the beta rollout of speech translation to consumer AI subscribers on Tuesday. The feature will initially be accessible in English and Spanish, with additional languages, such as Italian, German, and Portuguese, to be added in the coming weeks.
According to Google, the speech translation feature in Meet for Enterprises is currently being developed, with preliminary testing scheduled to commence with Workspace customers this year.
