GooglemodelGenie 4 (Research Preview)
Google DeepMind's new world model, which generates an interactive, explorable scene with sound that you can steer in plain language. Trusted testers only for now, with a waitlist.
Details
- You describe a change in plain language, like switching the weather, moving the camera, or adding something new, and the world updates in real time
- Gemini acts as the planner behind the scene: it turns your request into actions, plans multi-step events, and tracks what is happening as the story changes
- New in this version: ambient sound, music, and spoken dialogue with people in the scene, all generated together with the visuals
- All generated audio and video carries an invisible SynthID tag that marks it as AI-made
- Rolling out first as a research preview to trusted testers. Next up are Google AI Ultra subscribers in Project Genie and early-access partners in the Gemini API and Google AI Studio, with a waitlist open now
Best for
Prototyping a scene, a set, or a game level you can move through and change on the fly, instead of storyboarding it frame by frame.
When to use it
You need to show how a space or moment would look and sound before paying to shoot or build it, like VFX previs or a concept pitch.
Who benefits
Creative and video teams, game designers, and educators who pitch ideas visually.
Join the newsletter to get ahead with AI.
I share strategies, new modules, free guides, and tools to get you better at using AI in your work.
Leading a team? See team programs →