Google’s new AI transcription edits out your ‘ums’ and ‘ahs’
The Verge
Google has updated Gemini Audio with new 3.5 models that automatically transcribe speech, remove filler words, and handle background noise.Summary
Google has updated Gemini Audio with the new Gemini 3.5 models, introducing advanced transcription capabilities that include the Gemini 3.5 Transcribe model. This new tool automatically detects more than 85 languages, formats text, removes filler words like "um" and "uh", and allows users to edit naturally with their voice. Additionally, it supports custom vocabularies for specialized jargon, speaker attribution for up to three people, and word-level timestamps. These updates are rolling out across macOS, Android, and developer APIs.
(Source：The Verge)