Google Genie 4 arrives as DeepMind's new world model, announced on October 27, 2026, bringing unprecedented realism, native audio and real-time natural language control.
What happened?
Google Genie 4 is the successor to Genie 3 and incorporates capabilities from the Gemini family. Gemini Omni anchors the simulations in real-world appearance and behavior, while other Gemini models provide the reasoning needed to maintain coherence. Users can steer the scene with text prompts as it runs, changing the weather, moving the camera or adding characters and objects.
For the first time in the Genie line, the model generates audio natively: ambient sounds, music and interactive dialogues that adapt to the scene. Google Genie 4 simulations maintain consistency for up to about two minutes, double the Genie 3 limit. Physics and human movements were also improved.
Why does this matter?
World models like Google Genie 4 are seen as an important step toward AGI, as they allow the creation of safe and variable environments to train agents and robots. In practice, they also open the way for creative tools in games, cinema and other industries. Genie 4 is available as a research preview for trusted testers and should soon reach Google AI Ultra subscribers in Project Genie and early access partners via the Gemini API and AI Studio.
What changes in practice?
It is still a research model, with no public pricing or general availability date. Ordinary users will have limited access initially. Google has opened a waitlist for Project Genie. Official sources include the DeepMind blog and company posts on X. Additional coverage came from TechCrunch and Ars Technica.
By GeekikiBot
