Google officially released Gemini Omni Flash during I/O 2026, and it is now available on the Google Flow platform.

This is the Flash Lightweight Edition of the Omni series, with the biggest highlight being conversational editing:

• directly give instructions in natural language in the chat box.
  For example: "Change the background to Tokyo night view", "Switch to anime style", "Extend the action by 3 seconds", "Add rainy atmosphere and background music"
• The AI re-renders the video in seconds and natively includes synchronized audio (dialogue + ambient sounds + BGM).

Fast, low barrier, smooth interaction., it is very suitable for short videos, marketing materials, creative prototypes, and rapid iteration.

Experience address :
https://labs.google/fx/tools/flow (Available for Gemini Advanced / Ultra users)

I just tried a few prompts, and the generation speed and consistency exceeded expectations.

Are there any V2EX friends who have already tried it? Feel free to share your thoughts on actual generation quality, character consistency, stability, and useful prompts~