Edit video by transcript
Edit the words, not a timeline — cross out a sentence and that part of the video is cut.
The first run downloads the speech model to your browser (cached after that). Best on desktop — mobile devices may be slower.
Drop a video or audio file to edit by transcript
or click to browse
How it works
Drop a video or audio file in. WavyVid transcribes it with AI, right in your browser.
Click any sentence in the transcript to mark it for removal — it's struck through, not deleted, so you can undo anytime.
Export: the marked sentences are cut and the rest is stitched back together, with an updated caption file to match.
This is the text-based editing workflow popularized by tools like Descript, rebuilt as a free, private, browser-only tool. It's built for cutting rambling asides, false starts, and redundant sentences from talking-head videos, interviews, and voiceovers — anywhere the fastest edit is "just cut what I said there," not dragging clips on a timeline. Cuts happen at sentence boundaries (based on the AI transcript's own segmentation), not individual words.
Frequently asked questions
Does this cut at the exact word, or the whole sentence?
The whole transcript segment (typically a sentence or phrase, as detected by the AI transcription) — clicking a segment marks that entire chunk of speech for removal.
Can I undo a cut before exporting?
Yes — Undo/Redo work on every edit until you hit Export, which is the only irreversible step (it re-encodes the final video).
What happens to captions after I cut sentences?
You get an updated .srt alongside the edited video, with timestamps automatically corrected to match the new, shorter timeline.
Does it auto-detect filler words like "um" and "uh"?
Yes — likely filler words and stammered repeats are highlighted automatically, with a one-click "Remove all" you confirm before it's applied.