Google has announced Gemini 3.5 Live Translate, its latest audio model designed to make real-time conversations across different languages feel more natural. The new model brings speech-to-speech translation support for more than 70 languages and is being integrated across Google Translate, Google Meet, and developer tools.
The company says the new Gemini-powered translation model focuses not just on converting words from one language to another, but also on preserving elements such as tone, pacing, and the flow of a conversation.
Gemini 3.5 Live Translate: What's New
Unlike traditional translation tools that often wait for a speaker to finish before generating an output, Gemini 3.5 Live Translate is designed for continuous translation. This means users can have more natural back-and-forth conversations with reduced pauses between speakers.
Google says the model can understand speech, translate it, and generate translated audio while maintaining characteristics such as the speaker's delivery style. This could make translated conversations feel closer to speaking with someone who understands the same language.
Where Users Will See Gemini 3.5 Live Translate
The feature will be available across multiple Google products. In Google Translate, it is aimed at everyday use cases such as travel, conversations, and language assistance. Google Meet will use the model to improve live translated conversations during video calls. Developers will also be able to access the model through the Gemini API and Google AI Studio.
For Google Meet, the upgrade significantly expands language coverage, moving beyond limited language support to more than 70 languages.
Why This Matters
Real-time translation has existed for years, but the biggest challenge has been making translated conversations feel natural. Delays, robotic voices, and loss of emotion often make interactions feel less personal.
With Gemini 3.5 Live Translate, Google is focusing on making AI translation more conversational rather than simply more accurate. The move also highlights how AI models are moving deeper into everyday communication tools, where voice and context are becoming just as important as text-based responses.
