Google Announces Gemini 3.5 Live Translate for Real-Time Voice Translation
Google has introduced Gemini 3.5 Live Translate, a speech-to-speech model capable of instant translation across more than 70 languages. The system aims to preserve a speaker's tone and pitch while incorporating SynthID watermarks to identify AI-generated audio. The technology is rolling out across the Google ecosystem, including Google Meet, the Gemini Live API, and the Google Translate app.