Google announces Genie 4, a world model in a limited research rollout
Google's blog, dated 27 October 2026, announces Genie 4, a world model for real-time interactive simulation. The post says a person can set an event in motion and steer the scene with natural language, such as moving the camera, changing the weather or introducing a character, and that each world generates native audio, including ambient sound, music and interactive dialogue. It says the model builds on the Gemini family, with Gemini Omni anchoring appearance and behaviour and other Gemini models used for reasoning. Google says it has started rolling Genie 4 out to a limited group of trusted testers, and that it is coming soon to Google AI Ultra subscribers in Project Genie and to early access partners in the Gemini API and AI Studio. The post does not state a price or a context window.
Source: Google (opens in a new tab), 27 Oct 2026Also: Google DeepMind, 27 Oct 2026 (opens in a new tab)
Google's blog, dated 27 October 2026, announces Genie 4, a world model for real-time interactive simulation. The post says a person can set an event in motion and steer the scene with natural language, such as moving the camera, changing the weather or introducing a character, and that each world generates native audio, including ambient sound, music and interactive dialogue. It says the model builds on the Gemini family, with Gemini Omni anchoring appearance and behaviour and other Gemini models used for reasoning. Google says it has started rolling Genie 4 out to a limited group of trusted testers, and that it is coming soon to Google AI Ultra subscribers in Project Genie and to early access partners in the Gemini API and AI Studio. The post does not state a price or a context window.
The DeepMind post the same day calls the rollout a research preview for trusted testers, and says the company is gathering feedback before a release to consumers and developers, starting with those same Ultra subscribers and early access partners. It says Gemini 3.8 Flash serves as a guiding agent while Genie 4 generates a world, watching the scene and breaking a goal into steps. It says the audio in the accompanying demonstrations is produced by Genie 4.
What it means for business: Genie 4 is not generally available. A generated scene that shows a real person, or that includes speech, can raise likeness, consent and synthetic-content labelling questions once a wider release exists. The posts do not set a price.