Google DeepMind Unveils Genie 4 For Real-Time AI Worlds With Built-In Audio: All Details

Google DeepMind released Genie 4 in research preview, calling it its most realistic and interactive world model yet. Users type a text or image prompt and get a photorealistic world they can explore, and they can change it live while it runs.

What is new

Genie 4 is DeepMind's first real-time world model with built-in audio. Music, ambient sounds and interactive dialogue are generated together with the world. Users can type prompts mid-simulation to change the weather or camera angle, or to add a person or an object, while the scene keeps going.

DeepMind says the visuals are built on Gemini Omni and that the physics is better, with more convincing aerodynamics and collisions and more natural human movement. A scene stays consistent for up to two minutes, double the limit of Genie 3, which launched in August 2025.

The robotics link

In a research experiment tied to Gemini Robotics, a Gemini 4 Argon model broke tasks into steps that Genie 4 simulated in real time. The point is to rehearse robot scenarios in simulation before trying them in the real world. DeepMind also says worlds like these could train agents such as its SIMA game-playing system. Waymo adopted Genie 3 in February 2026 to build a specialised world model for autonomous-driving simulation, according to Wikipedia.

Intended uses

Reports list gaming, interactive learning and robotics as the main commercial uses. One speculative example is stepping into a painting of a historical scene and talking to the people in it.

Limits

Genie 4 isn't open to the public. A summary of TechCrunch's coverage says details can still drift over longer simulations and that audio and in-world text accuracy remain unsolved. Two minutes is also short for a game or a long training run.

Who can use it

Genie 4 is available now to trusted testers. In the coming weeks it will reach Google AI Ultra subscribers through Project Genie, Google's tool for creating interactive worlds from prompts, and early-access partners of the Gemini API and Google AI Studio. Google hasn't given a wider date or India availability in the reports I reviewed.

The competition

The launch comes months after AI video startup Runway released GWM Worlds 2, its second world model, which also responds to user inputs and generates audio, IT Home and NewsBytes reported. Alibaba has also announced a world model called Happy Oyster. World models have become a competitive area because the same technology applies to games, robots and self-driving cars.