Features

Everything it does, and what it deliberately does not

Each of these is a switch. You turn on what this video needs and leave the rest alone — nothing is applied without you turning it on.

The list below is the whole product. There is no hidden pro tier on this page.

Cutting pauses is free and needs no account.

How it works

This is what happens the moment you open a recording, before you decide anything.

2 pauses removed · 1.9 s shorter
Silence and pauses go first. This runs in your browser, on your machine, before anything is uploaded — and it is free.
9:16 output 16:9 source
The vertical crop follows the face. One framing per take, not per frame — a crop that chases the face frame by frame shakes, and shake reads as amateur.
TodayI'llshowyouhowthisworks
Captions come out of the speech itself, already in sync, with a screen to fix proper nouns before you export.
raw corrected
White balance, levels and contrast are measured on your own footage — and when there are two cameras, the app matches them to each other before anything else.

What you get

Every one of these is a switch. Turn on what this video needs and leave the rest alone.

Pauses cut automatically

Silence, hesitation and dead air are removed in the browser, before anything is uploaded. Free, no account.

The AI reads what was said

False starts, repeated lines and off-script chatter are flagged one by one, for you to approve or keep.

Captions from the speech

Already in sync, with ready-made styles and a screen to fix proper nouns before export.

Reframing that follows the face

16:9, 9:16, 4:5 and 1:1 from one recording. The crop tracks whoever is speaking, one framing per take.

Colour measured on your footage

White balance, levels and contrast from your own material, plus looks. Two cameras are matched to each other first.

Voice levelled and cleaned

Noise reduced, every scene brought to the same loudness, peaks held under the ceiling you choose.

Soundtrack that ducks under you

Generated to the length of your edit, dropping on its own while you speak.

Person separated from background

Blur it or push it back so the speaker comes forward. The desktop app uses a heavier model for this.

Name bar, highlight and end card

Lower third, one sentence in focus and a closing card with your logo — placed on the preview itself.

Hand it to a real editor

Export the timeline as XML/EDL/SRT for Premiere, Final Cut and DaVinci — with the reframing, a .cube of the grade, and the person matte.

Your video never leaves

The file is read straight from your computer. Only the audio goes out, and only when you ask for captions or take selection.

A desktop app for long material

No duration ceiling, a heavier background model and a stronger colour pass — the render writes to disk.

And what it does not do

Stated here so you find out now and not after you have recorded.

Up to 30 minutes per project

In the browser. The ceiling is memory during export, not a plan — the desktop app has none.

Firefox writes WebM, not MP4

It opens on a computer and on YouTube, but Instagram and TikTok tend to refuse it. Export from Chrome, Edge or Safari for those.

Export needs a free account

Cutting pauses does not. Five videos a month while the beta lasts, because captions and take selection use a server.

No multi-track timeline

On purpose. This is for people who record, not for people who edit — if you need layers, export the timeline to Premiere or DaVinci.

Questions

The ones people ask before trying it.

What is NinjaClip?

NinjaClip is an AI-powered automatic video editor that turns raw footage into a polished, ready-to-publish video: it cuts the pauses, picks the good takes, writes the captions, reframes the speaker and corrects the colour. It runs in your browser, and there is also a desktop app for long material.

What does NinjaClip do?

It removes silence and hesitation, reads the transcript to flag false starts and repeated lines, generates captions in sync, reframes 16:9 footage to 9:16, 4:5 or 1:1 following the speaker, levels the voice, corrects colour, adds a soundtrack that ducks under speech, and exports the finished video or a timeline for Premiere, Final Cut and DaVinci.

Does NinjaClip work in the browser?

Yes. Editing runs entirely in the browser and your video file is read straight from your computer — it is never uploaded. Producing the final file works in Chrome, Edge, Safari and Firefox; Firefox writes WebM instead of MP4.

Is NinjaClip free?

Cutting pauses is free and needs no account. Captions, take selection, soundtrack and export need a free account, because they use a server — five videos a month while the beta lasts.

How long can the video be?

Up to 30 minutes of combined footage per project in the browser — the ceiling is the browser's memory during export, not a plan limit. The desktop app has no ceiling.

What files can I open?

Whatever your browser opens: MP4 and MOV in H.264 or HEVC, and WebM. You can drop several files at once and they become one video, in the order you picked them.

Does it work in languages other than English?

Yes. The spoken language is detected automatically, so the captions and the transcript come out in whatever was said. The interface is available in English and Portuguese.

Can I finish the edit in Premiere or DaVinci?

Yes. Export a package with the cut timeline as .xml, .fcpxml and .edl, captions as .srt, the reframing already applied as scale and position on each clip, and a .cube with the colour. The desktop app adds the person matte.

Can I give it the script?

Yes — paste the script or open a .txt, .md or .docx. The selection then judges by delivery: which take says each line best. If the script does not match the footage, the app says so and falls back to judging each take on its own.

Related pages

Same product, described for a different starting point.

Open a raw video and watch the first edit appear

Nothing to install, no account to try it, and your file never leaves your computer.

Start editing