New Delhi: Google has recently unveiled Gemini Omni 1.1 Flash, the latest upgrade to its generative video model that adds scene extension, first and last frame controls, 360p video generation, 1080p and 4K outputs, and support for video references. Gemini Omni 1.1 Flash can continue an existing video from where it left off. This model can analyze up to 10 seconds of prior video context, compared with the final second in earlier models.
This offers additional context for maintaining visual consistency and narrative continuity when expanding a scene. Videos can be expanded in 10-second increments, up to a 40-second cumulative length. This model enables users to specify the starting and ending frames of the shot and generates video between the two keyframes. It is going to support camera orbits, zoom transitions, looping clips, and other camera movements between two defined frames. Gemini Omni 1.1 Flash supports video generation at 360p resolution for previews and faster iteration.
The company said 360p generation is up to 60% faster and costs one-third as much as standard 720p generation. However, the 360p mode can be used for rapid prototyping, storyboard iteration, and quick rendering on developer platforms. Gemini Omni 1.1 Flash supports 1080p and 4K video outputs via upscaling, enabling generated videos to be produced at higher resolutions.
This model can accept up to three seconds of video reference as part of its multimodal input. Reference video can offer visual context and help maintain character consistency when generating a scene. Gemini Omni 1.1 Flash is rolling out via the Gemini API in Google AI Studio. Enterprises can also use Omni 1.1 Flash via the Gemini Enterprise Agent Platform API. The company is offering: Developer documentation, a cookbook, and prompting guides covering scene extension, video references, and upscaling.
