Gemini Omni 1.1 Flash lets you build with more control

Google has unveiled Gemini Omni 1.1 Flash, a production-ready update designed to provide developers with enhanced control over generative video creation. This new version presents several key features aimed at improving the video generation process. One such feature is the ability to extend scenes seamlessly by analyzing up to 10 seconds of prior context, a significant improvement over previous models that only considered the final second.

This allows for improved visual consistency and narrative adherence, enabling the creation of longer stories or exploration of new creative avenues.

Developers can now specify the start and end frames of shots to achieve smooth transitions and camera movements. This capability facilitates the generation of continuous video between two keyframes, making it particularly useful for complex camera orbits, zoom transitions, or looping clips. Additionally, the update allows for the generation of lightweight previews in 360p resolution, up to 60% faster and at a third of the cost compared to the standard 720p resolution.

This feature is particularly beneficial for rapid prototyping, storyboard iteration, and quick rendering in developer platforms.

The new version also supports generating high-resolution outputs in 1080p or 4K, ready for professional production. Developers can maintain visual context and character consistency by using up to three seconds of video references when crafting their scenes. Furthermore, Omni 1.1 Flash introduces the capability to replace a sequence of uploaded videos with provided characters, allowing for the performance of individual dances in a cohesive, continuous shot.

This feature demonstrates the model's potential in enabling developers to build custom tools and creative workflows that leverage generative video capabilities.