Google has launched Gemini Omni 1.1 Flash, introducing major upgrades to its artificial intelligence video generation and editing technology, including longer scene extensions, improved camera controls and output resolutions reaching 4K.
The model is now generally available to developers through the Gemini API and Google AI Studio after previously operating in preview.
One of its biggest improvements is Scene Extension, which allows creators to continue existing videos while maintaining characters, environments and narrative consistency. Omni 1.1 can analyse up to 10 seconds of previous video context, compared with only the final second in earlier versions.
Scene Extension Allows Videos Up to 40 Seconds
Creators can extend videos in 10-second increments to reach a cumulative length of 40 seconds. Moreover, they can introduce fresh prompts during each extension to alter dialogue, camera movement or story direction.
Google has also introduced first-and-last-frame controls, allowing users to provide two keyframes while the model generates the continuous sequence between them.
Additionally, Omni 1.1 can process up to three seconds of reference video to transfer movement and other visual information into newly generated scenes.
New Draft Mode Speeds Up Video Creation
Meanwhile, Google has introduced a 360p draft mode for rapid experimentation and storyboarding. The company says it can generate videos up to 60% faster than standard 720p output while costing roughly one-third as much.
Creators can subsequently produce or upscale selected videos to 1080p or 4K. The API supports 360p, 720p, 1080p and 4K, with 720p remaining the default.
Google charges developers approximately $0.10 per second for standard 720p generation through its paid Gemini API.
Furthermore, the technology is rolling out through Google Flow and the Gemini app for eligible subscribers worldwide.
Google plans to retire the previous gemini-omni-flash-preview API endpoint on September 30 as Omni 1.1 Flash becomes its production-ready replacement.
