Google Gemini 3.5 transcription automatically removes filler sounds and supports 85+ languages

Google has updated Gemini Audio with three new Gemini 3.5 models — Live, Live Experimental and Transcribe — that automatically strip filler sounds like "um" and "ah" from transcriptions. The models support over 85 languages, can detect specialised jargon, and handle background noise more effectively than before.

Similar stories

💻 Technology Google rolls out Gemini interactive simulations to all Workspace users Android Authority · 1d ago 💻 Technology Google's Gemini update lets AI image generators skip removing the visible watermark Gizmodo · 11d ago 💻 Technology Developer claims Google Gemini leaked unreleased game details from private Google Doc TechRadar · 20d ago