Gemini Omni 1.1 Flash adds tighter controls for AI video builders

Google has released Gemini Omni 1.1 Flash, a production-ready update aimed at developers building generative video, creative applications and media-editing workflows. In its Google DeepMind Blog announcement dated Aug. 27, 2026, the company positions the model as a more controllable option for professional video production.

More context for scene extensions

The model can continue an existing clip while analyzing up to 10 seconds of preceding footage, compared with just the final second in earlier models. That broader context is designed to improve visual consistency and narrative continuity when extending a scene.

Extensions are generated in 10-second increments, with a cumulative maximum of 40 seconds. Developers can use the capability to lengthen a story, change direction, or maintain a planned cinematic sequence across multiple generations.

Faster iteration, higher-resolution output

Omni 1.1 Flash also adds controls for specifying a video’s first and last frames, enabling smoother transitions and more deliberate camera movements. Google highlights examples including dolly-zooms, snap-zooms and 360-degree camera orbits around a subject.

For prototyping, developers can generate 360p previews before committing to higher-quality output. Once a concept is approved, the model can upscale the finished project to 4K, creating a workflow that separates inexpensive experimentation from final delivery.

The update is available through the Gemini API in Google AI Studio, with access also offered through the Gemini Enterprise Agent Platform. For teams building AI-native editing tools, the combination of temporal context, frame-level control and preview-to-final workflows could make generated video easier to integrate into repeatable production pipelines.

Source: Google DeepMind Blog