A Submagic alternative that renders in your browser

Let us get the useful part out of the way first, because most comparison pages bury it: Submagic does considerably more than we do. Captions in over a hundred languages, pulling multiple clips out of one long recording, B-roll, avatars, scheduling, team workspaces. It also removes silences, so if you came here expecting "they don't cut pauses, we do" — that is not the difference.

If you want a broad short-form suite, that is the honest recommendation and you can stop reading. What follows is the narrow case where a smaller tool is the better answer.

Difference one: the video does not have to leave your machine

Caption & Cut renders in your browser through WebCodecs. The cutting and the caption burn-in happen on your own hardware, which means the footage is never uploaded anywhere.

Being precise about the limit, because a half-true privacy claim is worse than none: the audio track does leave your device. It is extracted locally, roughly a megabyte per minute, and sent for transcription. The picture stays with you; the sound does not. If that distinction matters for your material — client work under NDA, medical or legal content, anything unreleased — it is the whole reason this tool exists in this shape.

Difference two: where the cut comes from

Every tool in this category transcribes your audio and gets back a word list with start and end times. The obvious next step is to cut on those timings. They look precise. They are not.

Measured against the waveform on our own test clips, they were off by up to 0.9 seconds. In one recording the word "Captions" ends at 23.30s according to the transcript and is actually spoken until 24.20. Cut on the transcript and you slice a word in half at full volume. If you have ever heard an automatic edit clip the end off a word, this is why.

Caption & Cut takes its cut points from a loudness measurement of your original audio instead, and uses the transcript only for the caption text. That is a design decision with a cost — it means we cannot cut inside a sentence as aggressively — and a benefit, which is that word endings survive.

Difference three: you see every cut before it happens

Before anything renders, the transcript appears with every planned cut struck through: the pauses, the filler words, the takes you started over. One click puts any of them back.

This matters more than detection quality. A tool that silently removes a sentence that mattered has ruined the video, and you will only find out on playback. In our own tests the language model occasionally wanted to cut a passage that was explaining something to the viewer, purely because it mentioned the recording. That is exactly the kind of decision you want to see rather than discover.

Side by side

Feature list for Submagic taken from their own site, September 2026. We have deliberately left their pricing blank rather than quote a number that may be out of date — check it at the source.

What Caption & Cut will not do

It is built for talking-head material: explainers, podcast cuts, tutorials. It is not a music editor and not a multi-camera tool, it does not tell speakers apart, and it will not rescue clipped audio. There is no B-roll and no avatar, and there is not going to be.

Try it

Three videos free, nothing to install, runs in the browser. Free results carry a small watermark. Upload a video.

Related: remove silence from a video · captioning without uploading · all features