Google has launched Gemini 3.5 Live Translate, a new AI model enabling real-time voice-to-voice translation in over 70 languages with improved latency and natural-sounding speech. The technology is being integrated across Google's ecosystem, including Google Meet, and is available to developers through Gemini Live API and AI Studio. This represents a significant advancement in making real-time translation more accessible without requiring specialized hardware.
Background
Google has been developing real-time translation technology for years, previously requiring specific hardware like Google phones or earbuds. The company has been gradually expanding access through its Translate app before this latest advancement.
- Source
- Ars Technica
- Published
- Jun 10, 2026 at 02:57 AM
- Score
- 8.0 / 10