Remove filler words from your video automatically

"Um", "uh", "like", "so": every unedited recording is full of them. Finding and cutting each occurrence by hand can easily cost half a day on a 30-minute video.

The problem with manual editing

Filler words only stand out when you listen closely, and that is exactly what makes cutting them so tedious: you have to listen through the whole recording, mark every spot, and nudge the timeline frame by frame. On longer recordings this work grows in step with the runtime while your concentration drops with every minute. Skip or rush this step, and you publish a video that sounds choppier than what was actually said deserves. Do it properly, and you lose the time that was meant for your next video instead.

How AgenticCutter solves it

AgenticCutter transcribes your recording locally with mlx-whisper and detects filler words directly in the text, not in the waveform. The cutter proposes the cuts against the transcript, you read along and decide: keep it, cut it tighter, or hold a take, entirely in plain language via Claude Code or Codex. Only once you approve does the cutter render. The Cut-Grade slider also lets you dial how aggressively it cuts, from gentle to uncompromising. Your footage stays on your Mac throughout, there is no video upload. Especially for solo creators who publish regularly and know their own footage best, whether talking-head or screencast, this is the single step that saves the most time per episode.

Related features

• Remove filler words and ums
• Adjustable cut aggressiveness
• Transcript-driven editing

Related use case: Generate subtitles