Google launches Gemini 3.5 Transcribe with jargon detection and 85+ language support

The Verge AI · rss · 2026-08-27

Google updated Gemini Audio with new Gemini 3.5 models, introducing transcription capabilities that automatically detect specialized jargon and support over 85 languages. Designed to improve precision for Google's voice-controlled AI, the models maintain performance despite background noise or speech interruptions. The new Gemini 3.5 Transcribe model is a key addition to the family.

Original post →

More from Models

Models channel →