What happened
Google has quietly rolled out one of the most business-relevant updates to its AI video generation stack this year: Gemini Omni 1.1 Flash. The new version of Google's video model brings a faster, cheaper draft mode, longer scene extension, and more creative control over camera movement — all changes aimed squarely at teams that need to produce video content at scale without burning through budget.
The headline number is a 360p draft mode that renders up to 60% faster and costs roughly a third of the price of 720p generation. On the pricing side, Gemini Omni 1.1 Flash now charges $0.03 per second at 360p, $0.10 per second at 720p, $0.15 per second at 1080p, and $0.30 per second at 4K. That's a meaningfully wider range than the original Gemini Omni Flash, which only offered a flat $0.10-per-second rate at 720p.
### Longer, more consistent scenes
Scene extension has also been overhauled. Instead of analyzing just the last second of an existing clip to predict what comes next, Omni 1.1 now reviews up to ten seconds of prior footage before extending a scene. That context window produces noticeably more consistent motion, lighting, and character continuity. Scenes can now be stretched in 10-second increments up to a maximum of 40 seconds, up from much shorter continuations previously.
Developers also gain two new creative controls: uploading up to three seconds of external footage as a style reference — useful for carrying a recurring character, brand look, or motion pattern across multiple generated clips — and setting explicit start and end frames to script camera movement between two keyframes.
Why it matters
For any business producing video at volume — social ads, product demos, explainer content, localized marketing — cost per second has always been the limiting factor in AI video generation. A 60% speed boost combined with a two-thirds price cut on draft renders changes the economics of iteration. Teams can now generate five or six 360p drafts for roughly the cost of one previous-generation clip, test different hooks or scripts, and only upscale the winning version to 1080p or 4K for final delivery.
The extended scene length (up to 40 seconds) also matters for anyone building narrative or product-demo content, where a single continuous shot beyond a few seconds was previously hard to achieve without visible drift in character appearance or background detail. Combined with the ten-second lookback window, this update specifically targets the visual-consistency problem that has been the biggest complaint about AI-generated video so far.
How to use it today
Gemini Omni 1.1 Flash is live now through Google AI Studio and the developer API documentation, so there's no waitlist to access the new pricing tiers or scene-extension features. Marketers and creators who want to prototype ideas without touching code can start in 360p draft mode, iterate on concept and pacing at low cost, then move to the API or AI Studio's export options once a version is locked in for upscaling to 1080p or 4K.
For teams that want to test AI video concepts before committing to a paid API workflow, free resources like the tools at [mykreatool.com](https://mykreatool.com) are a practical starting point — useful for quick scripting, storyboarding, or asset prep before feeding a concept into a paid model like Omni 1.1 for final rendering.
### Comparing the options
Google now runs three distinct video tiers side by side: Gemini Omni Flash, Gemini Omni 1.1 Flash, and the Veo 3.1 family (Lite and Fast). Veo 3.1 Lite is actually cheaper at 720p ($0.05) and 1080p ($0.08) than Omni 1.1, but it has no 360p draft tier and no 4K option. Veo 3.1 Fast matches Omni 1.1's 4K price ($0.30) but costs more at 1080p. In practice, Omni 1.1 is the only tier offering a full 360p-to-4K pricing ladder, which makes it the most flexible choice for teams that want to scale spend up or down based on the project.
Who benefits
Small marketing teams and solo creators stand to gain the most from the new draft-mode economics — 360p iteration at $0.03 per second means testing ten concept variations costs less than a single finished 720p clip did before. Agencies producing client video at volume benefit from the longer 40-second scene extension, since it reduces the manual stitching previously needed to build longer continuous shots. E-commerce brands and app developers benefit from the style-reference upload, which makes it easier to keep a mascot, spokesperson, or product consistently rendered across a batch of ad variations.
Risks
As with any generative video tool, output quality at the cheapest tier (360p) is a draft-grade result, not a finished asset — businesses still need to budget for upscaling costs on any clip they intend to publish. There's also a growing need for clear disclosure: as AI-generated video becomes cheaper and more consistent, audiences and platforms increasingly expect labeling of synthetic content, and brands that skip this risk trust and compliance issues, particularly in regulated industries like finance, healthcare, and political advertising. Finally, per-second pricing across four quality tiers adds complexity to cost forecasting — teams running high-volume campaigns should model total spend carefully before scaling up 4K exports.
Conclusion
Gemini Omni 1.1 Flash is a practical, cost-driven upgrade rather than a flashy one: faster and cheaper draft renders, longer and more consistent scene extension, and finer creative control over camera movement and style. For businesses already using AI video generation in their marketing or product workflows, the update lowers the cost of experimentation while raising the ceiling on what a single generated clip can do. It's available now through Google AI Studio, with no special access required to start testing.
Comments 0