Skip to content
News MLQ.ai Aug 2026

Google: Gemini Omni 1.1 Flash adds scene extension and 4K upscaling

Google released Gemini Omni 1.1 Flash on August 27, 2026, adding several production-focused capabilities to its generative video model. The most significant change is scene extension: the model can now analyze up to ten seconds of prior video context when generating a continuation, up from one second in the earlier version, which improves visual consistency and narrative flow across extended sequences. Individual generations remain 3–10 seconds, but segments can be chained into cumulative outputs up to 40 seconds.

Two other additions matter for design and creative workflows. First and last frame interpolation allows users to specify a starting and ending image, with the model generating the transition between them — useful for UI transitions, brand animations, and storyboard sequences where the key frames are already decided. A 360p draft mode reduces cost by roughly two-thirds compared to the 720p default, enabling faster and cheaper iteration cycles before committing to a high-quality render. Finished work can then be upscaled to 4K (the upscaling is algorithmic rather than native generation, which the coverage notes but which will be invisible to most viewers at typical output sizes).

Figma, Adobe, and Runway have embedded the model into their platforms as early integrators. For designers already working in Figma or Adobe tools, this means access to generative video generation without leaving existing workflows.

The model is available in Google AI Studio and the Gemini Enterprise Agent Platform for developers, with consumer access rolling out via Google Flow to AI Plus, Pro, and Ultra subscribers.