Gemini Omni 1.2 Flash: Google stretches AI video to 15 seconds and cuts the price by 30%
Google made Gemini Omni 1.2 Flash available on November 5, 2026: its model for generating and editing AI video now creates clips of up to 15 seconds instead of 10, costs 30% less than version 1.1 and lets you fix a video with follow-up requests instead of starting over. Here is what changes, where you can use it and what it really costs.
Leggi questo articolo in italiano
What Gemini Omni 1.2 Flash is and when it launched
Gemini Omni 1.2 Flash is Google's model for generating and editing video with audio from text, images and other references. Google introduced it on November 5, 2026 in a post by Anish Nangia, a product manager at Google DeepMind, and the next day the Google Cloud blog declared it generally available.
It is not a brand-new model: it is the update to Gemini Omni 1.1 Flash. "Flash" is the line built for speed and cost rather than maximum quality at any price. There are three changes: longer videos, a lower price and a more practical way to fix what is wrong.
If you want an overview of the Omni family before the details, see our Gemini Omni guide.
What's new: 15-second clips, references and multi-turn editing
The most visible change is length: Gemini Omni 1.2 Flash generates and edits clips of up to 15 seconds, up from 10 in version 1.1, and you set the length by the second, from 3 to 15. Here is what Google lists:
| Feature | What it does |
|---|---|
| Length | 3 to 15 seconds, in one-second steps |
| Resolution | Draft at 360p or 720p, then upscale to 1080p or 4K |
| References | Character, product or style images, short motion clips (up to 3 seconds) and audio cues |
| Multi-turn editing | Swap product, background, lighting, weather or aspect ratio with follow-up requests |
| Continuity | First and last keyframes, plus clip extension |
| Speech | More natural dialogue, cleaner lip-sync, several speakers with distinct voices |
Multi-turn editing is the change that matters most in real work: instead of regenerating the whole clip when one detail is wrong, you ask for the fix and, according to Google, the model keeps everything else as it was. Google also claims fewer unwanted jump cuts in complex camera moves.
Two things to read carefully. First, on Google Cloud, audio references are allowlist only. Second, Unite.AI reports Google DeepMind's internal tests, in which 1.2 wins 62% of visual-consistency comparisons against 1.1; these are Google's own figures, not independently verified.
How much Gemini Omni 1.2 Flash costs
Google's stated price is $0.07 per second of 720p video, 30% less than Gemini Omni 1.1 Flash. Google Cloud says other resolutions see "similar savings" but does not publish the figures in the announcement.
Based on Gemini API pricing, Unite.AI works out about $0.70 for a 10-second clip and about $1.04 for a 15-second clip. The model is not on the Gemini API free tier: you need a paid account.
If you already call Gemini Omni 1.1 Flash through the API: according to Unite.AI, from December 3, 2026 requests to 1.1 are routed automatically to 1.2 at the new rates. Retest your prompts before then, because results can change.
Where to use it: Gemini API, Flow, Google Vids and the Gemini app
Gemini Omni 1.2 Flash is rolling out across several Google products, each with its own rules:
- Gemini API and Google AI Studio: for developers, with the model ID
gemini-omni-1.2-flash. - Google Cloud: for businesses, from the console or the API, including reserved capacity (Provisioned Throughput).
- Flow: Google's tool for directing scenes and editing in several passes. New to it? Read our Google Flow guide.
- Google Vids: 15-second clips "at no additional cost" with a Google account or a Workspace subscription. More in our Google Vids guide.
- Gemini app: 15-second videos rolling out to Google AI Ultra subscribers.
- Google Ads: to produce video for campaigns.
The announcements do not mention country restrictions. Every generated video carries SynthID, the invisible watermark Google uses to mark content made by its AI.
What it means for you if you make AI video
Fifteen seconds is the typical length of a hook plus one scene, or of a short social ad: with Gemini Omni 1.2 Flash you can finish a full shot in one generation instead of stitching two 10-second clips. At roughly a dollar per clip, redoing a scene three or four times stays affordable even on a small project.
Multi-turn editing changes the workflow: draft at 720p, fix things with precise requests ("change only the background", "same scene at night") and upscale to 1080p or 4K only the take that works. If you want to learn how to build scenes, prompts and edits with video models, start with AI Filmaking or the quick course Create stunning AI videos in 5 minutes. If you would rather have us make the video, there is our 30-second AI video service.
To compare Omni with the alternatives released in recent weeks, read our articles on Pika H3 Flex and Kling 4.0. And if you want a complete path that brings together video, images, voice and selling, there is The AI Creator System.
Frequently asked questions
When was Gemini Omni 1.2 Flash released?
Google announced it on November 5, 2026; on November 6, 2026 Google Cloud confirmed it as generally available.
How long are Gemini Omni 1.2 Flash videos?
Up to 15 seconds per clip, up from 10 in version 1.1. You choose the length by the second, from 3 to 15.
How much does Gemini Omni 1.2 Flash cost?
Google states $0.07 per second for 720p video, 30% less than version 1.1, so a 15-second clip costs about $1.04. There is no free tier for this model in the Gemini API.
Can I use Gemini Omni 1.2 Flash for free?
According to Google, in Google Vids you can generate 15-second clips at no additional cost with a Google account or a Workspace subscription. In the Gemini app, 15-second videos are for Google AI Ultra subscribers.
Can Gemini Omni videos be identified as AI?
Yes: Google applies the invisible SynthID watermark to every video, which lets it be identified as content generated with Google's AI.
Read next


