Google releases Gemini Omni 1.1 Flash with tighter controls for AI video tools

Google released Gemini Omni 1.1 Flash on August 27, positioning the model as a production ready update for developers building generative video tools. The company says the release is available through the Gemini API in Google AI Studio, with enterprise access through the Gemini Enterprise Agent Platform.

The change is less about one new demo and more about control. Google says Omni 1.1 Flash can extend an existing clip by reading up to 10 seconds of prior context, compared with the final second in earlier models. Extensions can be generated in 10 second increments, up to a cumulative 40 seconds. For teams making prototypes, ads, training media, game concept reels or creator tools, that context window matters because continuity failures are often where AI video stops being useful.

What Developers Get

The release adds first and last frame controls, letting a developer provide the starting and ending images for a shot and have Omni generate motion between them. Google frames this as a way to create camera orbits, zoom transitions and loops with more predictable results than prompt only generation.

Google also added a lower resolution draft path. The company says 360p previews can generate up to 60 percent faster and at one third of the cost of the standard 720p mode, based on its throughput comparison. Finished projects can then be rendered at 1080p or upscaled to 4K.

Why It Matters

The practical value is workflow separation. Creative teams often need cheap iterations before committing to a final render. A faster preview tier gives product builders a clearer way to design approval flows, storyboard tools and automated editing interfaces without charging every experiment at final output cost.

• Developers can test Omni 1.1 Flash directly in Google AI Studio.
• Enterprises can build with it through the Agent Platform API.
• Scene extension, frame interpolation and high resolution output are the core new controls.

CyberOGZ sees the key tradeoff as verification rather than access. Better video controls make AI media tools more useful, but they also make generated footage easier to polish. Builders should pair these capabilities with content labels, review steps and storage of prompts or provenance data. The strongest products will not just generate cleaner clips. They will make it obvious how a clip was created, approved and changed before publication.

Sources

• blog.google

Cover photo by Amar Preciado on Pexels, used under the Pexels License.