Ten seconds is a strange length for a video. It is enough for a reaction, but not for a scene: an establishing shot, an action, and a reaction never quite fit. Google's AI video model has just fixed that. On November 5, 2026, Google released Gemini Omni 1.2 Flash, and the headline change is simple — clips can now run up to 15 seconds long, with better consistency, conversational editing, and a price cut of 30 percent. I am Muhammad Hafeez, and in this guide I will explain what changed, where you can actually use the longer clips, what they cost, and how to make your first 15-second video. If you enjoy short AI clips, my guide on how to make a GIF in ChatGPT covers a similar idea with a different tool.
Table of Contents
What Is Gemini Omni 1.2 Flash?
Gemini Omni 1.2 Flash is Google's video generation and editing model — the tool that turns a text prompt, a photo, or an existing clip into a short video with sound. It is the successor to Gemini Omni 1.1 Flash, and Google made it generally available on November 5, 2026 across the Gemini API, Gemini for business, Flow, Google Vids, and the Gemini app.
The name is confusing, so let me clear it up. "Gemini Omni" is Google's video model family. It is not the same as Gemini 4, the text and reasoning model family — if you want that story, read my Gemini 4 explained guide. Omni makes videos; Gemini 4 answers questions and writes.
Three things changed in version 1.2, according to Google's launch post:
- Longer clips. You can now generate and edit videos natively up to 15 seconds long, up from 10 seconds with Omni 1.1, choosing any length from 3 to 15 seconds.
- Multi-turn conversational editing. You can fix a clip by talking to the model — "make the lighting warmer" or "slow down the camera move" — instead of starting over from a blank prompt.
- Stronger consistency. Faces, products, and characters stay more consistent across a clip, which was the most common complaint about the earlier version.
Where You Can Use It Today
The model is generally available, but the 15-second limit is not the same everywhere:
| Place | What you get | Cost |
|---|---|---|
| Google Flow | Text, frames, and references to video; multi-turn edits; 15 seconds, Extend, and upscaling need a Google AI Plus, Pro, or Ultra plan | Free accounts stay at 10 seconds; credits are charged per second |
| Google Vids | 15-second clips inside the Vids video editor | No additional cost with a Google account or Google Workspace subscription |
| Gemini app | 15-second videos | Rolling out to Google AI Ultra subscribers |
| Gemini API / Google AI Studio | Full control, 3 to 15 seconds, model ID gemini-omni-1.2-flash | About $0.07 per second at 720p |
For most beginners, the practical choice is Flow or Vids. Flow is Google's dedicated AI filmmaking tool, and it is where the editing features live. Vids is the simpler option: if you already make videos in Google Vids, the 15-second clips cost nothing extra.
How to Make Your First 15-Second Clip in Flow
Flow is the best place to learn the model, because it shows the prompt, the frames, and the editing tools in one place. Here is the basic workflow:
Step 1: Open Flow and Start a New Project
Go to Flow with your Google account and start a new project. Describe your video in the prompt box. If you have a Google AI plan, the 15-second option will be available; on a free account, clips stay at 10 seconds.
Step 2: Write a Prompt That Fits 15 Seconds
A 15-second clip needs a prompt with a beginning, a middle, and an end. The extra five seconds buy you one more beat — an establishing shot, then the action, then a reaction. Keep it to one scene; the model still works best with a single location and a single action.
Step 3: Set the Duration
Choose any length between 3 and 15 seconds. Start with 15 only if your prompt has enough happening to fill it — a static scene stretched to 15 seconds will feel slow.
Step 4: Edit With a Follow-Up Message
This is the new part. If the first result is close but not right, do not start over. Send a follow-up message in plain language — "make the lighting warmer" or "slow down the camera move" — and the model revises the same clip. Multi-turn editing is the biggest practical improvement in 1.2, because regenerating from scratch was the old workflow's biggest waste of time and credits.
Step 5: Extend and Upscale If You Need Them
Flow's Extend feature adds 3 to 10 seconds per turn, up to a 30-second maximum. The old Veo Extend was retired on November 5, 2026, so extending now requires Omni Flash 1.2 on a paid plan. Upscaling is also a paid-plan feature.
How to Make a 15-Second Clip Free in Google Vids
If you do not want to pay for Flow credits, Google Vids is the free route. Vids is Google's video editor, and the November 5 rollout added 15-second Omni clips at no additional cost for anyone with a Google account or a Google Workspace subscription.
Open Vids, start a video, and use the AI video option to generate a clip. The clips work the same way — text prompt in, short video out — but inside an editor where you can add more clips, titles, and voiceover around them. For a short social video or a school project, Vids is the simpler tool: generate the clip, drop it on the timeline, and keep building.
The Limits You Should Know Before You Start
The 15-second headline comes with fine print, and knowing it now saves wasted credits:
- First and last frame interpolation stays at 3 to 10 seconds. If you use the first-and-last-frame mode, asking for more than 10 seconds returns an error. The 15-second limit applies to standard generation and editing.
- Free Flow accounts stay at 10 seconds. The 15-second option, Extend, and upscaling all need a Google AI Plus, Pro, or Ultra plan in Flow. A 10-second 720p clip in Flow now costs 30 credits, where it cost 15 on Omni 1.1, and Google says credit costs can change at any time — check the cost in Flow before you generate.
- Reference videos shrink for long clips. For 3-to-10-second outputs you can use up to 3 reference clips; for 11-to-15-second outputs, only 1, and each clip can run up to 3 seconds.
- Audio references are still limited. Audio-reference input remains allowlisted rather than generally open, so most users cannot feed in their own audio yet.
- The Gemini app is Ultra-only for now. 15-second videos in the Gemini app are rolling out to Google AI Ultra subscribers, not to everyone.
- Some regions are excluded. Video editing and avatar creation are unavailable in the European Economic Area, Switzerland, the United Kingdom, and some US states, according to Google.
- Old models are being retired. From December 3, 2026, requests to Omni 1.1 will automatically route to 1.2 and be billed at 1.2 rates. Veo 3.1 Fast, Veo 3.1 Quality, and Omni Flash 1.1 are retired in Flow on December 5, 2026.
There is one more thing every creator should know: all generated videos carry an imperceptible SynthID watermark, Google's system for marking AI-generated content. You cannot remove it, and you should not try to. It is there so viewers and platforms can tell the video was made by AI.
Common Mistakes
These are the mistakes to avoid with Gemini Omni 1.2 Flash:
- Writing a 10-second prompt and asking for 15 seconds. A prompt with one action stretched to 15 seconds looks slow and empty. Give the model three beats — setup, action, reaction — or keep the clip shorter.
- Paying for 15 seconds in Flow on a free account. Check your plan first: free accounts stay at 10 seconds, and the 15-second option needs a paid Google AI plan. The free way to get 15 seconds is Google Vids.
- Regenerating from scratch instead of editing. Multi-turn editing is the new feature — use a follow-up message to fix a clip. Starting over burns credits and loses the parts that were already right.
- Expecting a finished commercial from one clip. Fifteen seconds is enough for a complete scene, but continuity, audio, rights, and factual accuracy still need human review. One generation is a raw clip, not a finished video.
- Forgetting the SynthID watermark. Every clip is watermarked as AI-generated. If you publish it, be honest about how it was made — the same disclosure rule applies as with any AI content.
- Using first-and-last-frame mode past 10 seconds. That mode caps at 10 seconds. Asking for more returns an error, so plan your workflow around the 10-second limit if you need frame interpolation.
Frequently Asked Questions
What is Gemini Omni 1.2 Flash?
Google's video generation and editing model, released November 5, 2026. It creates and edits clips from text, images, and video, with native audio, in clips of 3 to 15 seconds. It is a different model from Gemini 4, which handles text and reasoning.
How long can a Gemini Omni 1.2 Flash video be?
A single clip runs 3 to 15 seconds, in one-second steps. Extending adds 3 to 10 seconds per turn, up to 30 seconds. First-and-last-frame workflows are limited to 3 to 10 seconds.
How much does Gemini Omni 1.2 Flash cost?
Through the API, about $0.07 per second at 720p, which is 30 percent less than Omni 1.1 — so a 15-second clip costs around $1.05. In Flow, you pay with credits per second, and a 10-second 720p clip now costs 30 credits. In Google Vids, 15-second clips cost nothing extra with a Google account or Workspace subscription.
Can I make 15-second videos free?
Yes, in Google Vids — 15-second clips are included at no additional cost with a Google account. In Flow, 15 seconds needs a Google AI Plus, Pro, or Ultra plan; free accounts stay at 10 seconds. In the Gemini app, 15-second videos are rolling out to AI Ultra subscribers only.
What is the difference between Gemini Omni 1.2 Flash and Gemini 4?
Omni is Google's video model — it makes and edits video clips. Gemini 4 is the text and reasoning model family, with Argon for the public and Carbon for cyber defenders. They share the Gemini brand but do different jobs, and the guides for each are separate.
Conclusion
Gemini Omni 1.2 Flash is a practical upgrade, not a revolution: five more seconds per clip, editing by conversation instead of starting over, better consistency, and a lower price. The five seconds matter more than they sound — they turn a reaction into a scene, with room for a setup, an action, and a reaction in a single generation.
The smartest way to try it is free: open Google Vids, generate a 15-second clip with a three-beat prompt, and see what the model does with it. If you like the result and want more control, Flow is the next step — just check your plan and the credit cost first, because 15 seconds, Extend, and upscaling are paid features there.
0 Comments