Genie 4 is Google DeepMind's world model: an AI that simulates a place you can move through in real time, makes its sounds, and changes it when you type.
In plain words
A world model imagines how a place looks and what happens next in it. In Genie 4 you move and turn the camera with the keyboard, and type prompts to change the weather or add a person or object. It also generates music, background noise and dialogue along with the scene. It builds on Gemini models; Gemini Omni keeps scenes looking and behaving like the real world. DeepMind says details can still drift as a session runs longer. Trusted testers get it first, in a research preview.
A quick example
Imagine walking through a simulated mountain village and typing "make it snow". The scene changes around you as you keep moving, and you hear the wind pick up.
Where you'll meet it
- Training environments for robots and AI agents.
- Turning a game idea into a prototype you can explore.
- Trying out film scenes before shooting them.
Related
Sources
- DocsGenie 4: A new generation of world simulation, Google DeepMind · read 27 Oct 2026
- DocsGenie, Google DeepMind · read 27 Oct 2026
Checked by WhatIsAI against the sources above · 27 Oct 2026. Spotted an error? Report it.