Google announced Gemini Omni Flash at I/O 2026. It’s now live on Google Flow.
The interesting part is conversational video editing. You can type instructions like:
change the background to Tokyo at night switch to anime style extend the scene by 3 seconds add rain and background music It re-renders the video with synced audio, including dialogue, ambient sound, and BGM.
Site: omniflashai.com
Tried a few prompts. Speed is better than expected, and the workflow feels smooth for short videos, marketing drafts, and quick creative prototypes.
Anyone else testing it? Curious about character consistency, output quality, and prompts that work well.