Google launches Gemini 3.5 Live Translate: real-time voice translation that keeps your tone

Google has released a new real-time voice translation model called Gemini 3.5 Live Translate. It can recognize over 70 languages and generate natural-sounding speech while keeping the speaker's tone, speed, and pitch.

Instead of waiting for a person to finish talking before translating, this model keeps generating speech continuously. It balances waiting for more context to improve accuracy and translating quickly to keep up with the speaker. There are no awkward pauses, and the translation is just a few seconds behind. Developers can try it through the Gemini Live API and Google AI Studio, while business customers can use it in Google Meet starting this month.

Regular users can use it through Google Translate on Android and iOS. Just connect any headphones to get smoother translations in over 70 languages. Android users will also get a "listen mode" where they can hold the phone to their ear like a phone call to hear the translation without headphones.