Never edited one before? This is the whole thing, in order: what to record, what to click, and what to check before you upload. Nothing here costs money, and the editing part happens on your own computer.
In a hurry? Open the editor and press Generate demo footage. You can walk all five steps with a working project before you record anything of your own.
Everything runs inside the browser tab, so nothing is installed and nothing is uploaded. Chrome and Edge are fully tested; other browsers usually work too but aren't guaranteed 100%. Phones and tablets can open the editor but can't export.
As a file on your disk. A film, an episode, a music video, a trailer: the editor never looks at what it is, it only sees a picture and a sound.
A built-in laptop webcam is enough to start. Headphones matter more than the camera: they keep the video’s sound out of your microphone.
If you already have the file on your disk, use it. It is sharper than any screen recording, it is smaller, and it stays perfectly smooth, because the editor reads it straight from your drive.
If you can only stream it, record your screen instead: OBS Studio → Display Capture for the picture and Desktop Audio for the sound. Expect a bigger file and slightly softer image, and remember that a platform’s terms still apply to you.
Any recorder works. OBS Studio is free and does the job in five clicks:
Wear headphones. Through speakers, your microphone records the video too. You get an echo that cannot be removed, and the automatic cutting fires in the wrong places.
Two files are enough to start: the editor uses the sound inside each recording as it is. But if your camera file also carries the video's sound, the editor hears both at once. It still cuts, just less precisely, and you can never turn the video down under your voice.
So when you can, give it your microphone as its own file. Either record it as a separate track, or pull it out afterwards with VLC, which is free and already on most machines:
microphone.mp3, and press Start.video-audio.mp3.Honest note: yes, this is five extra minutes, and it is the one part of the process that is not yet inside the editor. If your recorder can already write a separate audio file, skip straight past this step.
Open the editor. There are two cards: your camera and the video you react to. Drag one recording onto each; the editor reads length, resolution and the sound wave right there in the page. If a file carries its own sound, that is all it needs. Each card also has a discreet line for adding a separate sound file when you have one.
Your camera started before the video did. That is normal, and this step fixes it. You get both sound waves, one above the other. Drag until the same loud moment sits above itself, then press play and listen.
Two decisions, both with a live preview. First, who owns the screen: the video, with you in a corner shape, or you, with the video in a window. Then, optionally, a second look for the moments you talk, which is what makes a reaction edit feel edited.
This is the one part that takes real time, because your own machine is doing work that normally happens on someone's server. Before it starts you are told how much disk space it needs and whether you have it. A ten to fifteen minute project at 1080p usually takes a few minutes.
One generation = one finished video = one credit, and only if it finishes. The editor builds each look as its own working file so the next step can cut between them without re-encoding. That is why you see more than one file, not because you are paying more.
Leave the tab open. You can work in other windows, but closing this one throws the video away.
Running out of space? The browser only lets a site use a slice of your disk, often far less than you actually have free. If that is too small, the editor lets you choose a folder on any drive (another partition, an external disk) and writes the working files straight there, using that drive’s full free space.
The cuts are already placed when you get here. The editor read your microphone and switched look every time you started and stopped talking. Review them first: click a clip to jump to it, drag an edge, delete one you do not want. Undo goes all the way back.
If that is all you wanted, you are finished. Press save and the file is copied together rather than re-encoded, so it takes seconds and loses nothing.
But this screen is a full timeline, so nothing stops you carrying on. Titles, shapes, your own music and images each get a lane above the main track, any cut can take a transition, and colour is set per clip. None of it costs another credit. Everything the editor can do →
The reaction is done at step 8, but you are not locked into our cut. Every generation also hands you the three full-length files it was built from, each the whole project in one look, and a timeline (.edl). All of it downloads from the last screen, and none of it has to be re-rendered.
The look files are ordinary MP4s. Drop watching.mp4 and talking.mp4 into DaVinci Resolve, Premiere, CapCut, anything at all, and build a different edit at full quality.
Load the three files, then import the timeline. Export it as EDL, FCPXML, OTIO or CSV. Your NLE drops every automatically-placed cut back onto a track, frame-accurate (30 fps), ready to trim, reorder or restyle.
Even if you upload final.mp4 today, keep the look files: they are the whole project, so you can re-cut a different version months later without touching your original footage.
Every one of these is easy to avoid before you record, and painful after.
If your voice and the video’s sound are baked into a single track, they cannot be separated later: no automatic cutting, and any echo is permanent. Keep them apart from the start.
Your microphone will pick up the video, you will hear yourself twice in the export, and the speech detection will cut in the wrong places.
Let it run. Trimming is one drag in the editor; a missing thirty seconds is a re-shoot.
The work happens in that tab. Closing it throws the video away, and you can watch the progress instead: it tells you how long is left.
Working files are big. The editor shows the exact number before it starts; if it is red, clear some space first.
Chrome and Edge are tested from start to finish. Others often work but aren't guaranteed. Firefox, for instance, writes Opus audio inside the MP4 that the default Windows player refuses to open. If in doubt, export from Chrome or Edge.
Reaction videos can pick up Content ID claims, and it mostly comes down to how much of the original you show. The editor gives you levers to show less without losing your reaction. Here is how they map to real choices.
These reduce the chance of a claim; they do not remove it. Content ID is unpredictable, and reaction videos rely on fair use or fair dealing, judged case by case. You are responsible for what you publish, and nothing here is legal advice. Read the terms, and if a channel or real money depends on it, talk to a lawyer in your country.
The first one takes an evening because everything is new. The second one takes an hour, and every one after that opens with your layout already set.
Open the editor, free