Gemini 3.5 Live Translate is a new audio model delivering speech-to-speech translation across 70+ languages in near real-time.
- •Automatically detects languages and preserves speakers' intonation, pacing, and pitch in translated speech
- •Generates continuous speech output, staying only seconds behind the speaker instead of waiting for pauses
- •Rolling out via Gemini Live API (developer preview), Google Meet (enterprise private preview), and Google Translate apps (Android/iOS)
- •Processes multilingual inputs without manual configuration and operates reliably in loud, unpredictable environments
- •All generated audio is watermarked with SynthID to remain detectable as AI-generated content
This summary was automatically generated by AI based on the original article and may not be fully accurate.
Deepmind
The Hacker News
Hugging Face
Google Cloud