- Which filler words does it remove?
- The built-in list covers non-word vocalizations: um, uh, uhm, erm, er, ah, hmm, mhm and variants. Meaningful-but-overused words like "like" or "basically" are NOT cut by default — add them yourself in the extra-words field if you want them gone, since cutting them can change meaning.
- Will it clip the words around a filler?
- No — each cut is clamped against the neighboring words' timestamps with a small guard zone, so the edit points always land in the silence around the filler, not inside real speech.
- How much shorter does my recording get?
- Typical unscripted speech contains 3-8 fillers per minute plus pauses; a 20-minute podcast episode commonly loses 1-3 minutes. The result page shows exactly how many fillers were cut and the total time saved.
- Does it work on video?
- Yes — the video is cut at the same points as the audio so everything stays in sync. Video output is re-encoded (cuts require it) at high quality; audio-only files come back as MP3.
- Is this like Descript's filler removal?
- Same idea — automatic, transcript-driven filler removal — but as a simple browser tool: upload, process, download. No project setup, no subscription required to try it.
