News in Short
- Google DeepMind has announced Genie 4, a world model for real-time interactive simulation.
- Users can change scenes with natural-language prompts, including camera movement, weather and new characters.
- The model generates ambient audio, music and interactive dialogue as simulations unfold.
- Genie 4 is rolling out to trusted testers, with access planned through Project Genie and developer tools.
Google DeepMind has announced Genie 4, a new AI world model designed to create realistic, interactive simulations. The system lets users shape an environment while it runs, rather than simply generating a video to watch.
Users can set an event in motion and guide the scene with natural language. They can move the camera, change the weather or introduce new characters. Google says Genie 4 also creates audio in real time, including ambient sounds, music and interactive dialogue.
How Genie 4 builds interactive worlds
Genie 4 combines capabilities from Google’s Gemini family. Gemini Omni helps anchor simulations in real-world appearance and behaviour. Other Gemini models provide reasoning to help keep scenes coherent as users make changes.
The model can respond to prompts while a simulation continues. This allows users to explore a scene and adjust it on the fly. Google says its world model also simulates how people move, react and speak within the environment.
Native audio is another major feature. The model can generate environmental sounds that match a setting, alongside music and dialogue that reacts to user input or events. Google says this makes Genie 4 its first real-time world model with native audio generation.
Why world models matter
Google views world simulation as an important research step toward artificial general intelligence, or AGI. The company says these environments could provide safer training settings for robots and AI agents.
World models could also support creative and industrial work. Filmmakers may use them to explore scenes and camera angles. Game developers could test interactive concepts before building them with traditional tools. Google also points to potential applications in journalism and manufacturing.
Google’s work builds on its broader Gemini research. ITmatters has also covered Gemini’s personalized AI image generation, another example of how the company is expanding creative AI capabilities.
Genie 4 availability
Google has started rolling out Genie 4 to a limited group of trusted testers. The company says it is coming soon to Google AI Ultra subscribers through Project Genie.
Early access is also planned for partners using the Gemini API and Google AI Studio. Google has not announced a wider public release date. For now, Genie 4 remains an experimental research preview.
Google DeepMind says the model is still evolving. Its team is gathering feedback from testers before expanding access. Readers can explore the official Genie 4 announcement and demonstration videos for more details.
