Google Rolls Out Gemini 3.5 Live Translate for Real-Time Multilingual Conversations

Google on Wednesday rolled out its latest speech-to-speech translation model called Gemini 3.5 Live Translate. Google claims it is designed to enable more natural multilingual conversations. As per the company, the new AI model can detect more than 70 languages and generate translated speech. The result is claimed to preserve a speaker's tone, pacing, and intonation while continuously delivering near real-time translations. Gemini 3.5 Live Translate is rolling out across Google Translate, Google Meet, Google AI Studio, and the Gemini Live API for developers.

Gemini 3.5 Live Translate Processes Audio As It is Streamed in Real-Time

In a blog post, Google emphasised that Gemini 3.5 Live Translate is the next step in its translation efforts, which began nearly two decades ago as one of the company's earliest machine learning experiments. The company claims it is suitable for multilingual meetings, live broadcasts, lessons, customer support interactions, and real-time interpretation.

Gemini 3.5 Live Translate handles live translation with minimal delay, according to Google, and it is said to process audio as it is being streamed in real-time, generating translated audio a few seconds behind the original speaker. The company said it also avoids the pauses that are often associated with traditional turn-based translation systems.

The new speech-to-speech translation model can identify and switch between more than 70 languages, according to Google. It also eliminates the need to manually configure language settings. Gemini 3.5 Live Translate is said to remain effective in noisy and unpredictable environments.

Upgraded translation capabilities are also coming to Google Meet, the company announced. Like Live Translate, the app will support translation in more than 70 languages, which is a significant increase from the previous limit of five languages. Apart from this, the tech giant is bringing Gemini 3.5 Live Translate to the Google Translate app on Android and iOS. It will be available through Live Translate, and the translated speech can be heard through any paired compatible headphones.

On Android, Google is rolling out a Listening Mode that allows translated audio to be played directly through a smartphone's earpiece.

Tackling concerns around AI-generated content, the company noted that all audio generated by Gemini 3.5 Live Translate is embedded with SynthID, the company's watermarking technology. It is claimed to be integrated directly into generated audio and is designed to help identify AI-generated content.

Developers can access Gemini 3.5 Live Translate in public preview through the Gemini Live API and Google AI Studio. Google highlighted integrations with platforms including Agora, Fishjam, LiveKit, Pipecat, and Vision Agents, to help them build and deploy voice translation applications more easily. The AI model is rolling out globally across supported Google products starting today.