A quote from Google's blog.
https://blog.google/intl/ko-kr/company-news/technology/gemini-live-3-5-translate-kr/
Twenty years ago, Google's translation service started as one of the pioneering machine learning experiments designed to transform the science of language into a special experience connecting people with each other. That experiment has now achieved remarkable progress to the point where over one trillion words are translated across Google's products every month for billions of users.
Today, Google is taking the next step by launching 'Gemini 3.5 Live Translate,' the latest audio model for real-time speech-to-speech translation.
This model automatically detects over 70 languages and generates smooth, natural-sounding translated speech while preserving the speaker's tone, speed, and pitch. Unlike turn-by-turn systems that wait until the speaker finishes speaking to respond, Gemini 3.5 Live Translate continuously generates speech by balancing between waiting to understand the context to improve translation quality and translating in real-time to match the speaker's conversation pace. It provides seamless audio without awkward pauses, continuing the translation just a few seconds behind the speaker as the conversation progresses.
Starting today, Gemini 3.5 Live Translate is rolling out progressively across Google products as follows.
For developers: Available as a public preview through Gemini Live API and Google AI Studio
For enterprises: Available as a private preview in Google Meet starting this month
For general users: Available through the Google Translate app on Android and iOS
[Abbreviated]
Experience Gemini 3.5 Live Translate in Video Conferences
Gemini 3.5 Live Translate will soon be applied to Google Meet's speech translation feature, allowing you to experience the following benefits.
Expanded language support from 5 languages to over 70 languages
While previously only English-centric translation was available, now conversations in over 2,000 language combinations are possible in a single meeting
Updated interface for instant access to speech translation
This update launches as a private preview for select Google Workspace for Business customers starting this month, with broader deployment planned for the end of this year.
Google Meet participants communicating seamlessly in English, Mandarin Chinese, and Swedish using the real-time speech translation feature
Using Gemini 3.5 Live Translate in Google Translate App on Android and iOS
This model is also rolling out to the Google Translate app on Android and iOS worldwide. When using the real-time translation feature, simply connect headphones or earbuds to experience smoother translation across over 70 languages while preserving the speaker's tone.
For Android users, a new 'Listen Mode' based on Gemini 3.5 Live Translate, which allows you to hear translations directly through the phone receiver, will also roll out progressively. Simply hold the phone to your ear as you would during a regular call, and the translated audio plays immediately. This new feature is useful when you want to quickly hear the translated content without headphones and without others hearing it.
[Remaining content omitted]
