Genie 4 is Google DeepMind's natively multimodal world model: instead of rendering a fixed clip, it builds a playable, physics-aware environment you can steer in real time with plain language.
The genie4 guide covers the announced features, how prompting for interactive worlds differs from prompting for video, a side-by-side comparison with Genie 3, and the realistic access picture for creators and researchers.
