Google has launched 'Gemini Omni 1.1 Flash', an upgraded generative AI video model that gives creators greater control, faster experimentation, and improved consistency.
The model is available through the Gemini API in Google AI Studio and is designed for developers building video-generation workflows, creative tools and media-editing applications.
Longer Video Scene Extensions
One of the major upgrades in Gemini Omni 1.1 Flash is its ability to analyse up to 10 seconds of previous video context when extending a scene. Earlier models relied on a much shorter portion of the preceding video, making continuity harder to maintain.
With the new model, users can extend generated videos in 10-second increments, with the total video length reaching up to 40 seconds.
The longer context is intended to improve visual consistency and help preserve the flow of characters, objects and scenes throughout an extended sequence.
First-and-Last Frame Control
Gemini Omni 1.1 Flash also introduces greater control over how a video begins and ends. Creators can provide both the first and final frames of a shot, allowing the model to generate the video that connects the two points.
This capability can be used for controlled camera movements, zoom effects, camera-orbit sequences and seamless video loops.
It gives creators more control over a shot's progression, rather than relying entirely on text prompts to determine how the scene develops.
Faster 360p Video Generation
For creators who want to test ideas quickly, Google has added 360p video generation to Gemini Omni 1.1 Flash. Google says 360p previews can be generated up to 60% faster and cost about one-third as much as standard 720p output.
The lower-resolution option is aimed at tasks such as storyboarding, prototyping and experimenting with different creative concepts before producing a higher-quality version.
Once the concept is finalised, the model can generate videos at higher resolutions, including 1080p and 4K, allowing creators to move from quick previews to final-quality content.
Video Reference Support
The model also supports video references as part of its multimodal inputs. Creators can provide reference footage to help the AI maintain visual details and motion characteristics across generated scenes.
Google supports reference videos of up to three seconds for this workflow. The feature can help maintain consistency in elements such as character appearance and movement when creating multiple shots.
Gemini Omni 1.1 Flash Pricing
Google's reported API pricing varies according to the output resolution:
For a 10-second video, this works out to approximately $0.30 at 360p, $1 at 720p, $1.50 at 1080p and $3 at 4K.
Gemini Omni 1.1 Flash is available to developers through Google AI Studio and the Gemini API. Google is also bringing the model to Google Flow for AI Plus, Pro and Ultra subscribers globally, expanding access to its latest video-generation capabilities.
