You can now edit a video by typing a sentence. No timeline. No keyframes. No expensive editing software. Google’s Gemini Omni — the video model that replaced Veo in the Gemini app on November 5, 2026 — turns editing into a conversation. Type “swap the background for a beach at sunset,” and a new clip comes back with the change done.
The big shift: video editing is no longer a skill you learn. It is a request you make. Describe what you want in plain English, and you get a finished 10-second clip, an edited video, or a moving ad from a still photo. Here is how it works — and where it still falls short.
What Is Gemini Omni (and What Happened to Veo)?
Gemini Omni is Google’s video generation and editing model, introduced at Google I/O in May 2026. On November 5, 2026, Google made Gemini Omni 1.2 Flash generally available, and inside the Gemini app Omni is now the model behind every generated clip — Veo 3.1’s API previews shut down on October 22, 2026.
The difference from Veo is simple: Omni edits. Veo generated a clip and that was the end of it. Omni keeps the conversation going — say “make the lighting golden hour,” and it revises the same clip instead of starting over.
Clips run 3 to 10 seconds in the Gemini app; 15-second clips are rolling out to Google AI Ultra subscribers first. Through the API, 1.2 Flash supports 3 to 15 seconds at 720p or 1080p, with sound included.
What You Need Before You Start
Check three things before you open a prompt.
A paid Google AI plan. Video generation in the Gemini app needs a Google AI Plus, Pro, or Ultra subscription — a free account does not get it. The exception is Google Vids: any Google account can try it free at vids.new, with a small monthly allowance of AI clips.
You must be 18 or older. Google requires an 18+ account for video generation and editing in the Gemini app, Flow, and Vids.
Your location matters. Editing or extending a video you uploaded is not available in the European Economic Area, Switzerland, or the United Kingdom — and in the US, not in Illinois or Texas. There you can only edit clips the AI itself generated.
One more thing: every clip carries SynthID, Google’s invisible watermark marking it as AI-generated. Downloaded clips may also show a visible watermark.
How to Edit Video with Gemini: Step by Step
There are three ways to work with video in the Gemini app. Open gemini.google.com or the mobile app and sign in with your paid Google AI account.
1. Create a clip from a text prompt
- Tap the video option (“Create video”).
- Describe the clip in one sentence: subject, action, setting. Example: “A golden retriever puppy runs through a sunny park, slow motion, warm afternoon light.”
- Wait for the 10-second clip to render.
- Reply with one change at a time — “Make the light golden hour” — and Omni revises it.
- Download it when you are happy.
2. Turn photos into video
- Upload up to five photos.
- Describe the motion: “Animate this photo so the coffee steams and the camera slowly zooms in.”
- Gemini blends your photos into a moving clip with sound.
- Refine it in the same chat, one change per message.
3. Edit a video you already shot
- Upload one video clip (plus reference images).
- Describe the change: “Swap the background for a city skyline at dusk.”
- Gemini returns a new version with your change applied.
- Fine-tune by chatting: “Keep everything else the same, just brighten the sky.”
Two limits: edits work on up to 10 seconds of an uploaded clip, so plan around its best 10 seconds. And if a clip ends too soon, ask what happens next — Omni can extend the scene.
A Real Example: Product Photo to 10-Second Ad
An Ohio candle seller on Etsy has one good phone photo: a candle on a wooden table.
She uploads it to Gemini and types: “Animate this candle photo — the flame flickers, steam rises, the camera slowly moves closer. Warm, cozy evening light.”
Seconds later she has a 10-second clip with sound. She adds a title card — “Hand-poured in Ohio” — and posts it as an Instagram Reel.
The whole job took about five minutes and cost nothing beyond her Google AI subscription. A year ago, the same clip meant hiring a videographer or learning an editing app. That is the promise of Gemini Omni: short, polished video a non-editor can make between other tasks.
4 Prompt Tricks That Get Better Results
Say one change per message. “Change the background, the jacket, and make it rain” gets a muddled result. One change at a time keeps revisions clean.
Name the subject, the motion, and the light. “A barista pours latte art, close-up, soft morning light” beats “make a coffee video.”
Add “keep everything else the same” to small fixes. Without it, Omni sometimes reimagines the whole scene.
Plan around 10 seconds. Build longer videos as several 10-second scenes stitched together in Google Vids or any basic editor.
What Gemini Omni Can’t Do (Yet)
It is not a timeline editor. Gemini returns a new clip; it does not give you a timeline to cut, rearrange, and caption your own clips. For that, use Google Vids or a traditional editor. Vids is Google’s own answer: a browser-based editor with a real timeline, up to 30-minute videos, and 1080p MP4 export.
Clips are short. Ten seconds is the standard length — enough for ads, intros, and social clips, not a full YouTube video.
Regional limits apply. Editing your own uploaded video is blocked in the EEA, Switzerland, the UK, Illinois, and Texas.
It can wobble on details. Hands, small text, and fast motion sometimes come out wrong. Check every frame before you post.
The watermark is permanent. SynthID is embedded in every output, so Omni clips are always identifiable as AI-made.
Gemini App vs Google Vids vs the API
| Option | Best for | What you get |
|---|---|---|
| Gemini app | Quick clips, chat editing | 3–10s clips (15s on Ultra), photo-to-video, Video to Video Edit; needs a paid Google AI plan |
| Google Vids | Longer videos, work presentations | Scene timeline, transcript editing, up to 30 min, 1080p export; free basic access at vids.new |
| Gemini API | Developers, automation | Omni 1.2 Flash at $0.07 per second of 720p video; 3–15s clips; conversational edits |
The takeaway: use the Gemini app for fast single clips, Google Vids when you need a timeline, and the API when you are building video into a product.
Comparing AI tools more broadly? Our Gemini vs ChatGPT guide breaks down where each one wins, and our Gemini 4 Carbon guide covers Google’s newest model family. For the safety side, see is Gemini AI image generation safe.
Frequently Asked Questions
Can Gemini edit videos for free?
No — the Gemini app needs a paid Google AI plan (Plus, Pro, or Ultra) and an 18+ account. Google Vids is the free route: any Google account can try it at vids.new with a small monthly allowance, though Omni editing there needs a paid plan.
How long can a Gemini video be?
In the Gemini app, clips run 3 to 10 seconds, with 15-second clips rolling out to Google AI Ultra subscribers first. The API supports 3 to 15 seconds. Use scene extensions or Google Vids to build longer videos.
Can Gemini edit a video I recorded on my phone?
Yes. Upload one clip and describe the change — a new background, a wardrobe swap, a different style. This is Video to Video Edit. It works on up to 10 seconds of your clip, and it is not available in the EEA, Switzerland, the UK, Illinois, or Texas.
Does Gemini replace Veo?
Inside the Gemini app, yes — Omni is now the video model, and Veo 3.1’s API previews shut down in October 2026. Veo 3.1 still exists on Google Cloud, and Google Flow still offers both.
Do Gemini videos have a watermark?
Yes. Every clip carries SynthID, Google’s invisible watermark, and downloaded files may also show a visible one. You cannot remove it.
The Bottom Line
Video editing just became a conversation. Gemini Omni will not replace a professional editor — 10-second clips and no timeline see to that. But for the everyday jobs — a product ad from a photo, a new background for a social clip — it turns a half-day task into five minutes of typing. Open the Gemini app, upload one photo, and describe what should move. That first clip will tell you more than any guide can.