Google DeepMind Releases Genie 4 World Model With Real-Time Editing and Audio
By James Dargan · · 1 min read
Google DeepMind has released Genie 4, its newest world model, in research preview. The model produces its most photorealistic and interactive simulated worlds yet, letting users type prompts mid-simulation to change weather, camera angles or add people and objects, while generating matching real-time audio including music, ambient sound and dialogue.
Genie 4 maintains consistent scenes for up to two minutes, double that of Genie 3, with improved physics and more humanlike characters. Gemini is more tightly integrated, acting as a director that plans steps and checks frames. In a Gemini Robotics experiment, Gemini 4 Argon broke tasks into steps that Genie 4 simulated in real time, pointing to ways of testing robot scenarios before real-world deployment.
Research director Tejas Kulkarni said future world models will increasingly be hybrids, with Gemini writing code and text and Genie simulating reality in real time. Engineer Diego Rivas said trusted testers see value in film prototyping and storyboarding, though he stressed it is early days and noted object drift and audio accuracy remain challenges.
The release comes as the sector consolidates, with AMD acquiring Fei-Fei Li’s World Labs and Rivian buying Third Dimension. Genie 4 will reach Google AI Ultra subscribers and select Gemini API partners in the coming weeks.
