Google Gemini Omni 1.2 Flash: What’s New in 2026
Gemini Omni 1.2 Flash is Google’s newest AI video generation model, launched on November 5, 2026, and it makes short-form video creation longer, cheaper, and easier to edit. Google made the model generally available the same day it announced it. The headline changes: clips now run up to 15 seconds instead of 10, the price of 720p video dropped about 30 percent to roughly $0.07 per second, and editing works more like a conversation than a restart.
What Is Gemini Omni 1.2 Flash?
Gemini Omni 1.2 Flash is Google’s fast, cost-focused model for generating and editing video with AI, and the successor to Gemini Omni 1.1 Flash. The model ID is gemini-omni-1.2-flash. It is available through the Gemini API for developers, and inside Google products such as the Gemini app, Google Flow, and Google Vids.
What’s New in Gemini Omni 1.2 Flash
1. Clips up to 15 seconds long
One generation now runs up to 15 seconds, up from 10 seconds with Omni 1.1 Flash, and you can set the length in one-second steps from 3 to 15 seconds. Five extra seconds is a 50 percent increase — enough for a full product shot or a complete call to action without cutting the clip in half. In the Gemini app, 15-second generation is rolling out first to Google AI Ultra subscribers. In Google Flow, 15-second clips need a Google AI Plus, Pro, or Ultra plan. In Google Vids, they are included at no additional cost with a Google account or Workspace subscription.
2. A price cut of about 30 percent
Google cut the developer price by roughly 30 percent compared with Omni 1.1 Flash. On the Gemini API, video output now costs $12.00 per million tokens, down from $17.50, which Google converts to about $0.07 per second of 720p video. A 15-second clip at 720p therefore costs about $1.04 in video output — about $0.35 at 360p, about $1.56 at 1080p, and about $3.13 at 4K. Remember that this is the price of one generation pass, not of a finished video: drafts, retries, and revisions are billed too.
3. Better character and product consistency
The model reproduces reference images more faithfully, and it can work with up to six reference images at once. That is useful when a character or product has to look the same across several clips — a mascot in a series of ads, or a product shown from different angles. Consistency has been the weak point of AI video, so this change matters most for anyone building a series rather than a one-off clip.
4. Conversational editing
Instead of starting over when a clip is almost right, you can now edit it in turns. The model supports first-and-last-frame interpolation (3 to 10 seconds) and extending a clip by 3 to 10 seconds per turn, up to 30 seconds in total. You describe the change and the model revises the same scene.
5. Native audio included
Audio is generated by default and synchronized with the video, with no separate charge listed on the pricing page. Dialogue, sound effects, and ambient sound come from the same generation.
How Much Does It Cost?
Pricing is per token on the Gemini API, not per second — Google converts tokens into seconds for convenience. Text, image, and video input cost $1.50 per million tokens, and text output costs $9 per million tokens. Video output is $12.00 per million tokens; at 720p Google counts 5,792 tokens per second of video, which is how the $0.07-per-second figure is calculated. The model is paid tier only: the Gemini API pricing page lists no free tier for it.
Gemini Omni 1.2 Flash vs Omni 1.1 Flash
Maximum clip length rises from 10 to 15 seconds, with length selectable in one-second steps. Video output pricing falls from $17.50 to $12.00 per million tokens, about 30 percent cheaper — about $0.07 per second at 720p instead of about $0.10. Character consistency improves with up to six reference images, editing gains interpolation and extension, and Google doubled the paid-tier rate limits. From December 3, 2026, requests to the older gemini-omni-1.1-flash model route automatically to Omni 1.2 Flash and are billed at the lower 1.2 rates.
How to Get Started: A Simple 5-Step Plan
- Pick one small job. A 5-second product clip, a social teaser, or a storyboard scene — not a full advert on day one.
- Write a tight prompt. Describe the subject, the action, the setting, and the camera move in two or three sentences.
- Add a reference image. If a product or character must look right, upload a clear image of it and name it in the prompt.
- Generate short first. Start with a 3- to 5-second draft at 360p to test the idea cheaply, then extend the winner to 15 seconds at 720p.
- Edit, don’t regenerate. Use conversational editing and Extend to fix a clip instead of starting fresh — it is cheaper and keeps your scene consistent.
For context on the wider tool landscape, our roundup of AI video generator tools compares the other options worth knowing, and our guide to face-swap AI tools covers a related creative workflow. Our ranking of AI humanizers explains how to keep machine-written copy sounding natural.
Who Should Care — and Who Can Skip It
This update matters most for small businesses, freelancers, and creators who need short video regularly but cannot justify a shoot for every clip. At roughly $1 per 15-second 720p generation, a month of social content can cost less than a single stock-video license.
You can safely skip it for now if you need long-form video, precise lip-sync for dialogue-heavy scenes, or broadcast-grade quality. Fifteen seconds is still a short clip, and AI video still needs a human to check continuity, audio, and factual accuracy before anything is published.
Frequently Asked Questions
Is Gemini Omni 1.2 Flash free?
No. On the Gemini API it is paid tier only, with no free tier listed. Video output costs $12.00 per million tokens, about $0.07 per second at 720p. Inside Google products, 15-second clips in Google Vids carry no extra charge with a Google account or Workspace subscription, while in Flow they need a paid Google AI plan.
How long can a Gemini Omni 1.2 Flash video be?
A single generation runs from 3 to 15 seconds, in one-second steps. Using the Extend feature you can add 3 to 10 seconds per turn, up to 30 seconds in total.
When did Gemini Omni 1.2 Flash launch?
Google announced it and made it generally available on November 5, 2026. From December 3, 2026, requests to the older Omni 1.1 Flash model are routed to 1.2 automatically.
The Bottom Line
Gemini Omni 1.2 Flash is not a revolution — it is a practical upgrade that fixes the three things people complained about most: clips were too short, prices were too high, and editing meant starting over. Longer clips, a 30 percent price cut, better consistency, and conversational editing make it the best starting point for short AI video today.
