Google unveils Gemini Omni, the so-called "video version of Nano Banana," enabling video generation and editing through conversation

Google unveiled Gemini Omni at I/O 2026, a conversational AI model for generating and editing video with text, images, video, and audio inputs.

The first release, Gemini Omni Flash, is rolling out to eligible Gemini app and Google Flow subscribers, with YouTube and API support planned.

Google says Gemini Omni outputs will include SynthID and C2PA Content Credentials to help verify AI-generated video.

The launch extends Google’s generative AI into editable video while adding provenance tools for content authentication.