A Descript alternative that transcribes on your own machine
Strike a sentence in the transcript and the picture and the sound follow. The transcription runs in the tab (Whisper, in the browser), so the recording is never sent anywhere.
Editing by the words
- Transcription in the tab
- Whisper in the browser, two model sizes, many languages. Nothing is uploaded and nothing is metered.
- Speakers
- Told apart and named, so a two-hander reads as a script.
- Cut by striking text
- Delete a line in the transcript editor and the cut is made on the timeline, picture and sound together.
- Search what was said
- The bin finds a clip by the words spoken in it, and answers in timeline frames.
- Text on screen, too
- A scan reads the words in the picture, so a slide is found by its title.
- Captions
- A caption track with its own look, merged into sentences, and an .srt beside the export.
And a real timeline under it
Transcript editing is a way into the cut, not the whole of it. Underneath is an ordinary non-linear editor: ripple trim, nested sequences, keyframes on an anchor, 33 transitions, chroma key, LUT grading, a per-clip audio chain with a de-esser and a gate, auto-ducking, and loudness normalisation on the way out.
cutvid and Descript, side by side
| cutvid | Descript | |
|---|---|---|
| Price | Free | Free tier, then per seat |
| Transcription | In the tab, unmetered | On their servers, by the hour |
| Upload | Never | Required |
| Voice cloning / overdub | None | Yes |
| Filler-word removal | Silence strip; filler words by hand | One click |
| Team review | None | Yes |
| Timeline depth | Nests, keyframes, chroma key, LUTs | Scene-based, lighter |
| Works offline | Yes, once installed | No |
Where Descript wins
Overdub and voice cloning, one-click filler removal, studio sound, team review and comments, and a hosted project that follows an account between machines. None of that is here.
The rest of what cutvid does not do:
- Export is H.264 video and AAC audio in an MP4, or the sound alone as .m4a. No ProRes, no DNxHD, no H.265 out, no image sequences.
- No team features: no cloud project, no review link, no comments, no shared library. A project lives in one browser, and moves as a file.
- No stock library and nothing generative: no stock clips, no music bed, no AI b-roll, no generative fill.
- No third-party plugins. No OpenFX, no VST, no motion-graphics templates from another app.
- Colour is LUTs and the effect rail, not scopes, curves and secondaries. A .cube grades a clip; there is no vectorscope to grade against.
- Built and checked in Chromium browsers, which is where WebCodecs is furthest along. Other browsers vary by codec and by version.
- It runs in a tab, on one machine. A long 4K timeline is heavier here than in a native application on the same computer.
Questions
- Is the transcription free?
- Yes, and unmetered: it runs on the machine, so there are no minutes to buy.
- How accurate is it?
- It is Whisper, at one of two model sizes. The larger one is slower and better; both run in the tab.
- Is my recording uploaded for transcription?
- No. The model runs in the browser and the audio never leaves it.
- Can I edit the transcript text itself?
- The words of a recording are the recording’s: they are edited in the transcript editor, where the waveform and the speakers are. A caption written by hand on the timeline is typed over directly.
Open it and cut something.
No account, no install, nothing uploaded. The first edit takes as long as a page load.