How to Edit Videos With Gemini (Free, Step by Step, 2026)

Can Gemini edit videos?

Yes, but "edit" means two very different things, and most guides cover only one.

The first is Gemini Omni, Google's video model inside the Gemini app. You upload one short clip, describe a change ("make it rain", "swap the car for a bike") and Omni generates a new version. It changes what's in the shot. It doesn't cut a wedding, assemble a reel from twelve clips or caption a talking-head video.

The second is Gemini as the brain of a real video editor. Connect Gemini to a timeline editor and it reads your project, places cuts, adds titles, grades colour, styles captions and checks its own frames, while you watch and fix things by hand. That's what most people mean when they type "edit video with Gemini".

The guides ranking for this search today, like Curious Refuge's and Picsart's, focus on Omni, not on cutting several clips, adding text, exporting a file or doing it for free. Here's both, with the limits of each.

Way 1: generative edits with Gemini Omni (short clips, paid)

Use this when you want to change what a shot looks like, not how a video is cut.

1. Open the Gemini app (web, phone, or the Windows and Mac desktop apps) with a personal Google account on a Google AI plan.
2. Upload one clip. Third-party guides put the limit at about 10 seconds (Curious Refuge); Google's help page doesn't publish a number.
3. Describe the change and what must stay the same: "Add heavy rain. Keep the people, the lighting and the camera move exactly as they are."
4. Wait. Google says it "may take a few minutes", and you can't use that chat while it renders.
5. Keep editing the result in the same conversation, one change per turn.

Google's prompt guide lists what Omni can do: remove or replace objects and characters, change the camera angle or move, change the action, restyle the clip (anime, watercolour, claymation), add effects, and extend a clip in 10-second steps up to 40 seconds.

Where Google's help page draws the line:

• It needs a paid Google AI plan (or a qualifying Workspace licence). There's no free tier for editing in the Gemini app.
• Uploading videos to edit isn't available in the EEA, Switzerland, the UK and some US states, or to anyone under 18.
• It works on one clip at a time. No timeline, music track, captions or export settings.
• The output is a new generated video, not your original pixels. Faces, logos and text can drift between versions.

If you're in the UK or EU and the upload button isn't there, that's the regional block, not a bug. Way 2 below works everywhere.

Way 2: let Gemini cut your real footage (free)

For this you need an editor Gemini can control. We'll use Takejam, a browser editor built so any AI (Gemini, Claude or ChatGPT) can edit the same timeline you edit by hand. The Free plan covers everything below, and a Gemini key from Google is free too.

What you need

• Chrome or Edge on a Windows, Mac or Linux computer. There's no phone editing.
• A free Takejam account at app.takejam.com.
• A free Gemini API key from Google AI Studio. Google's pricing page lists the Flash models as free of charge.
• Your clips, music and images on your computer. Gemini can't upload files, and your media never leaves your machine.

Step by step

1. Sign in at app.takejam.com and create a project. Pick a size such as 9:16 for Reels or 16:9 for YouTube. You can change it later, or ask Gemini to.
2. Drag your videos, music and images in. Rename anything called IMG_4031 to something Gemini understands, like "interview-maria" or "drone-beach".
3. Click the AI tab on the left. Choose Gemini, paste your key into Your Gemini API key and click Use Gemini. Takejam checks the key and loads your models.
4. Keep the suggested Flash model for quick edits. Pick a Pro model from the dropdown for a big, multi-step edit.
5. Type what you want ("Ask Gemini to edit…") and press Enter. You'll see what Gemini is doing: "Reading the timeline", "Editing: Cut the intro to the beat", "Looking at the video".
6. Watch the timeline change. Don't like it? Click Undo AI change at the top of the editor, or ask Gemini to undo its last edit.
7. Finish by hand. Nudge a cut, move a keyframe, change a colour. Everything Gemini made is an ordinary clip on the timeline.
8. Click Export. The Free plan exports 1080p MP4, MOV or WebM with no watermark.

Your key stays in your browser and goes only to Google, never to Takejam. Google bills usage to the key, which on the free tier costs nothing within Google's limits.

Or connect Takejam inside the Gemini app

If you'd rather chat in Gemini itself, Google now lets you add your own apps. It calls them custom apps, and they work through MCP, the same standard Claude and ChatGPT use (see What is MCP?).

1. Go to gemini.google.com. At the bottom, click Settings → Connected Apps (if you don't see it, click Personal Intelligence first).
2. Under Custom apps, paste https://api.takejam.com/mcp and click Next.
3. A Takejam page opens. Sign in, choose the workspace Gemini may edit and click Allow.
4. In any chat, type @ and pick Takejam, then ask for your edit. Gemini asks you to confirm before it changes anything.

The catch: Google's help page says custom apps are only for people 18 or over in the US, on a personal Google account, in English, with Keep Activity on. Work and school accounts can't add them yet. Google's release notes say the new Customize menu for connecting apps began rolling out globally on 13 October 2026, so expect this to widen. Until then, the API-key route works in every country.

What should you ask Gemini to do?

Be specific about length, audience and feel. These all work in Takejam today:

For the beat-synced reel, detect the beat grid on your music track in Takejam first; Gemini then snaps cuts to it. For captions, generate them in the Captions tab (speech is transcribed on your computer, see our captions guide), then let Gemini style them.

How does Gemini check its own work?

Takejam gives Gemini a tool that renders frames exactly as the export would and hands them back as pictures: up to six chosen moments, or a contact sheet of up to 24 frames. So "look at 2, 5 and 9 seconds and fix anything odd" makes it render, look and correct before it says it's done. Flash models do this quickly; Pro models are slower but spot more.

Omni or a connected editor: which should you use?

They also combine well. Generate or restyle a hero shot with Omni, download it, drop it into Takejam, and let Gemini cut it into the rest of your edit.

What can't Gemini do in Takejam?

• Upload or find files. You add media; Gemini only uses what's in the project bin.
• Generate new footage. That's Omni's job, in the Gemini app.
• Watch your raw video. It sees a timeline description and a few small check frames, never your files.
• Transcribe speech. You click Generate captions yourself; Gemini can then restyle them.
• Change approved scenes. Scenes a reviewer marked approved are locked to the AI.

One honest note on privacy: Google's pricing page says content sent on the free API tier may be used to improve Google's products. In Takejam that's the timeline description and the preview frames, not your original files. For client work, use a paid Gemini key, which Google says is not used for training.

Is it free?

Takejam's Free plan includes the full editor, 1080p export with no watermark and 50 AI edits a day. Creator ($5 a month or $48 a year) makes AI edits unlimited and adds 4K, HEVC/AV1 export, the brand kit and all templates. Team ($8 per seat a month, minimum two seats) adds live co-editing, roles and client review links. The Gemini key is free from Google AI Studio within Google's limits; Omni edits need Google's paid plan.

Start at app.takejam.com, or read how the same setup works with ChatGPT and Claude.

Questions, answered