Filler removal
Remove filler words from a video or podcast in one click
So um, the honest answer is you know, we shipped it before it was ready uh, twice.
Rescript strips filler words from video and audio in one click, on your own machine, for free. It transcribes locally with Whisper, detects disfluencies like "um" and "uh" against that transcript, and cuts the matching audio and footage — leaving every cut visible and reversible. Nothing is uploaded and there's no per-minute charge.
- Filler word removal
- One click across the whole file
- Silence removal
- One click, pauses of 0.3s and over
- Price
- Free and open source for noncommercial use
- Where your media goes
- Nowhere. It never leaves your device
- Works offline
- Yes, once the model has downloaded a first time
- Source code
- Public on GitHub — auditable
Step by step
How it works, start to finish.
- 01
Load the recording
Drop your video or audio file into Rescript. Whisper transcribes it on your device with word-level timestamps — no upload, and no account.
- 02
Click Remove fillers
Every detected filler across the whole file is cut at once. Because detection runs against a word-level transcript rather than the waveform, the cut lands on the filler and not on the syllable next to it.
- 03
Review and adjust
Each removal shows as a struck-through word in the transcript and a red region on the timeline. Restore any word you want back, or drag a cut's edges if it's a fraction tight. Playback skips the cuts live so you can hear the result immediately.
- 04
Take out the dead air too
Remove silences cuts every pause of 0.3 seconds or longer in the same way. Between the two, most of the slack comes out of a recording before you've made a single manual edit.
Why word-level timing matters here
Filler removal is only as good as its timing. A hundred milliseconds early and you clip the previous word; late and the "um" is still audible as a stub. Tools working from audio energy alone have to guess at those boundaries.
Rescript cuts against word-level timestamps, so each filler has a start and an end of its own. Where the alignment is off, drag that word's edges on the timeline and the cut follows your correction.
Automatic, but not out of your hands
One click removes every filler in the file. But automatic removal gets things wrong — a deliberate pause, an "um" inside a quote, a misheard word — and a service that hands back a processed file gives you no way to know which.
In Rescript every cut is a visible, reversible region. The removed words stay struck through in the transcript where you can see exactly what went, and restoring one is a click. Nothing is decided off-screen.
Free, and nothing leaves the machine
Filler removal is metered by the hour on most services, which adds up quickly if you publish weekly. Rescript charges nothing for noncommercial use and has no hours cap, because the detection and the re-encode both run on your own hardware.
That also means the recording never gets copied anywhere. For interviews under NDA, research recordings, or anything with a participant who consented to one use and not to third-party processing, that's usually the deciding factor.
FAQ
Questions people actually ask
How do I remove filler words from a video for free?
Open the video in Rescript, let Whisper transcribe it on your device, and click Remove fillers. Every detected "um", "uh", and similar disfluency is cut across the whole file, and you can restore any of them. It's free for noncommercial use with no upload and no account.
Which filler words does it detect?
Common disfluencies — um, uh, er, and similar — in the transcript language. Anything it misses can be removed by selecting the word in the transcript and pressing delete, which cuts the matching footage the same way.
Does it remove breaths, mouth sounds, or stutters?
No. Rescript detects filler words and silences. Breaths, lip smacks, and repeated-word stutters aren't detected automatically — you can cut them manually on the timeline, but a dedicated audio-repair service will do more here.
Does it work on audio-only files?
Yes. Podcasts, voice notes, and interviews in MP3, WAV, or M4A get the same treatment, in an audio-only layout that drops the empty video pane and gives the transcript full width.
Will the cuts sound abrupt?
Cuts land on word boundaries from the transcript, which usually sounds natural. Where one is tight, drag either edge of the cut region on the timeline to loosen it, or double-click to reset it.
Keep exploring features
Transcript editing
Delete a word, the footage goes with it.
Silence removal
Cut every pause over 0.3s, then adjust any of them.
Speakers
Group the transcript by who is actually talking.
Timeline
Waveform, cut handles, and word-level timing by hand.
Regenerate
Rewrite a fumbled line, hear it in the original voice.
Weighing it against something else? See how Rescript compares.
Open it and see.
Free, open source, and running entirely on your own machine. No account, no upload, no watermark.
Or open the web app