Google DeepMind Releases Gemini Omni 1.1 Flash With New Video Tools

Google DeepMind has released Gemini Omni 1.1 Flash, the latest version in its Omni line of AI models, according to a blog post the company published on August 27, 2026.

The update adds new video-generation features, including the ability to extend an existing video scene, a first-and-last-frame interpolation tool that generates the video connecting two given frames, and the ability to upscale generated video to 4K resolution. DeepMind said the changes also allow faster prototyping.

DeepMind published example code showing that developers can call the new model, whose code name is gemini-omni-1.1-flash, through its Python software library to continue a previously generated video scene.