Top AI Product

Gemini Omni Flash is the first model in Google’s new Omni family — a natively multimodal generative media model that takes any mix of text, images, audio, and video as input and produces video as output. It shipped at I/O 2026 and is live today in the Gemini app, Google Flow, and YouTube Shorts.

## Conversational video, not a render queue

The pitch is short video generation and editing through plain conversation. Instead of a render-and-wait pipeline, you describe or adjust a clip in dialogue and iterate. Because the model is natively multimodal rather than a stack of stitched-together components, it treats images, audio, and video as one input space — so editing an existing clip and generating a new one are the same operation, not separate tools.

## Why it matters

Putting a conversational video model directly inside YouTube Shorts and the Gemini app is a distribution move as much as a technical one — it drops generative video into the apps where billions of people already make and watch short clips. The “Flash” naming signals the usual trade: speed and cost over maximum fidelity, aimed at everyday creation rather than studio work. That’s how generative video stops being a demo and starts being a feature people actually use.