Google has introduced a new AI-powered transcription feature that automatically removes filler words such as ums and ahs from audio recordings, according to a report by The Verge.
What Happened
The feature is integrated into Google's Gemini 3.5 model and is designed to clean up audio transcriptions by removing common speech disfluencies. Users can reportedly upload or record audio, and the AI will generate a transcription while automatically editing out filler words that speakers typically use when thinking or pausing.
Why It Matters
This development highlights the continued refinement of AI transcription tools beyond basic speech-to-text capabilities. By automatically cleaning up natural speech patterns, Google is targeting users who need polished transcripts for professional or content creation purposes. The feature adds to a growing list of AI tools designed to streamline audio and video production workflows.
The Bottom Line
Google's Gemini 3.5 now includes an AI transcription feature that removes filler words from audio recordings. The company reports the tool can generate cleaner transcripts suitable for professional use.