Google Launches Gemini 3.5 Transcribe, a Faster AI Speech-to-Text Model Supporting 85 Languages
Summary
Google launches Gemini 3.5 Transcribe, a powerful new AI speech-to-text model that is 70% faster than its predecessor, supports 85 languages, reduces live-speech error rates, and is already live in Gboard, the Gemini macOS app, and developer tools.
Key Points
- Google is launching Gemini 3.5 Transcribe, an AI-powered speech-to-text model that removes filler words, corrects speech on the fly, supports 85 languages, and handles up to three speakers in pre-recorded audio.
- The new model is approximately 70 percent faster than its predecessor Chirp 3 and reduces live-speech error rates from 7.32 percent to 5.5 percent, though it does alter wording to reflect intent rather than exact speech.
- Gemini 3.5 Transcribe is currently live in Gboard's Rambler feature on Pixel 11, the Gemini macOS app, and developer tools like AI Studio and the Gemini API, with a Chrome browser rollout coming soon.