Google DeepMind announces Genie 4, a world model with native audio

Google DeepMind announced Genie 4 on 27 October 2026 as its new world model for real-time interactive simulation. Users move through a generated world with standard keyboard controls and can change it with text prompts while it runs, for example to alter the weather, move the camera or add a person or object. Gemini adapts each prompt to the surrounding scene and keeps the world progressing consistently between inputs.

The model is natively multimodal: music, ambient sound and spoken dialogue are generated at the same time as the picture. DeepMind says Genie 4 builds on the Gemini family, with Gemini Omni grounding how scenes look and behave and other Gemini models providing the reasoning that keeps a simulation coherent. It also points to more realistic physics and human movement than before.

Access is limited for now. Genie 4 is rolling out as a research preview to trusted testers, and Google plans to offer it next to Google AI Ultra subscribers in Project Genie and to early access partners of the Gemini API and Google AI Studio. A waitlist is open. DeepMind presents world simulation as a step toward AGI and as a way to build training environments for robots and agents.