Google Gemini Omni 1.1 Flash Makes AI Video Creation More Precise and Flexible
Gemini Omni 1.1 Flash adds scene extension up to 40 seconds, first-and-last-frame interpolation, 360p drafts, and 4K output, with resolution-based pricing from $0.03 to $0.30 per second.
Google unveiled the video-generation and editing model Gemini Omni 1.1 Flash on August 27, 2026. Developers can use the model through the Gemini API and Google AI Studio. Key updates include extending scenes from existing videos, interpolation between start and end frames, low-resolution draft generation, and 4K upscaling.
Scene extension analyzes up to 10 seconds of the preceding video as context and continues the video in 10-second increments. The cumulative length is up to 40 seconds, and the model is designed to maintain characters and visual continuity for longer than the previous model, which referenced only the final second. By specifying start and end frames, users can also generate camera movement or transitions between the two keyframes.
The 360p draft mode is up to 60% faster than standard 720p and costs about one-third as much. Pricing based on the output video is $0.03 per second at 360p, $0.10 at 720p, $0.15 at 1080p, and $0.30 at 4K. Therefore, describing this model’s price simply as '$0.10 per second' is inaccurate without specifying the 720p condition.
Final results can be output at 1080p or 4K, and up to three seconds of external video can be provided as a reference to convey visual context for movement or characters. Adobe has integrated the existing Gemini Omni Flash into Firefly, while Omni 1.1 is also available for scene extension in Google Flow and the Gemini app.
This update is well suited to a production workflow that rapidly compares multiple 360p drafts and upscales only the selected result, rather than generating a finished version in a single pass. For developers building video-production tools, the biggest change is the ability to adjust generation cost, iteration speed, and scene control together.