Google DeepMind Releases Genie 4 World Model With Real-Time Audio

Google DeepMind has released Genie 4, its newest world model, in research preview for trusted testers. The company says it produces the most photorealistic and interactive worlds it has built so far, and it lets users type prompts mid-simulation to change what is happening on screen.

Genie 4 also generates audio in real time that matches the simulated world, including music, ambient sound and interactive dialogue. Freeform edits are possible too: users can change the weather or camera angle, or add new people and objects. A scene stays consistent for up to two minutes, double the limit of Genie 3, which launched as a research preview a little over a year ago. Physics and human-like characters have also improved.

Gemini as director

Gemini is now more tightly integrated. It acts as a director that plans steps and checks frames while Genie creates the visuals. In a research experiment tied to Gemini Robotics, a Gemini 4 Argon model broke tasks into steps that Genie 4 simulated in real time, pointing to a way of testing robot scenarios before they happen in the physical world.

"What we'll increasingly see in the future are these hybrid world models," said Tejas Kulkarni, research director at Google, describing Gemini writing code and text while Genie simulates reality at different levels of abstraction.

Google pitches gaming, interactive learning and robotics as the main uses. Trusted testers see another: prototyping films. DeepMind engineer Diego Rivas said real-time generation is most valuable for ideation and storyboarding. He added that users could eventually turn such prototypes into creative assets with Gemini Omni, though that is still some way off.

Limits and rollout

The team says it is early days and the end-to-end product experience is unclear. Details and objects can drift as a simulation runs longer, and the team has not yet worked out how to guarantee accurate audio or legible in-world text.

The release arrives as competition and consolidation grow in the field. Runway launched its second world model, GWM Worlds 2, a couple of months ago, while World Labs is being acquired by AMD, Amazon is in talks to buy Decart, and Rivian has acquired Third Dimension for self-driving simulation.

In the coming weeks, Genie 4 will reach Google AI Ultra subscribers in Project Genie, as well as early access partners of the Gemini API and Google AI Studio.