Google has quietly shifted the balance of power in generative video with the release of Gemini Omni 1.1 Flash, a model that treats scene extension not as a novelty but as a production-grade tool. The update allows creators to extend clips in ten-second increments up to a cumulative forty seconds while analyzing up to ten seconds of prior footage for context, a change that materially improves visual continuity and reduces the jarring inconsistencies that have plagued earlier iterations.
A new first-and-last-frame control lets directors specify two keyframes and have the model generate the motion between them, opening the door to controlled camera orbits, zoom transitions, and seamless looping shots. Alongside this, a draft mode renders at 360p resolution up to sixty percent faster and at roughly one-third the cost of standard 720p output, enabling rapid prototyping before committing to expensive high-resolution renders.
The model is already live for developers through Google AI Studio and the Gemini Enterprise Agent Platform, with consumer access rolling out to Google AI Plus, Pro, and Ultra subscribers via Google Flow. Major creative platforms are moving quickly to integrate the technology, with Adobe bringing it into Firefly and design teams at Figma describing it as a shift from merely generating videos to actively directing them within their workflows.
Pricing remains opaque for the highest resolutions, but early API rates suggest video output costs around seventeen dollars and fifty cents per million tokens, or roughly ten cents per second of 720p footage. With scene extension also arriving in the Gemini app and the ability to supply up to three seconds of existing video as a multimodal reference, Google is positioning Omni 1.1 Flash as the backbone of a new, more controllable era of AI-driven storytelling.
#GeminiOmni #AIVideo #GenerativeAI #GoogleAI #VideoProduction #TechNews