Google Enhances AI Video Generation with Gemini Omni 1.1 Flash

Google is significantly upgrading its AI video generation capabilities with the launch of Gemini Omni 1.1 Flash. This new model moves beyond simple text-to-video generation, offering users much greater control over the creative process and targeting professional video production needs.

The key feature of Gemini Omni 1.1 Flash is its emphasis on "controllability." Users can now extend existing video clips, with each extension up to 10 seconds, referencing the previous 10 seconds to ensure a natural continuation of characters, scenes, and narrative. This allows for the creation of videos up to 40 seconds long.

Furthermore, users can specify the first and last frames of a scene, enabling the AI to generate the intermediate steps. This is particularly useful for creating smooth transitions, zoom effects, or looping videos. The model also supports the use of reference videos, allowing up to 3 seconds of footage to guide the AI in maintaining visual consistency and character coherence.

For faster iteration and cost-effectiveness, Gemini Omni 1.1 Flash includes a 360p preview option that generates footage up to 60% faster and at one-third the cost of 720p. The final output can be rendered in high resolutions, including 1080p and upscaled 4K, suitable for professional editing. Pricing for API usage varies by resolution, starting at $0.03 per second for 360p video.