Google DeepMind Launches Genie 4 World Model With Audio
Google DeepMind has released Genie 4, introducing real-time audio generation, improved physics, and longer simulation runtimes.
Radwan AbdulhadiEditor and technology writer
Introduction to Genie 4
Google DeepMind has released its newest artificial intelligence world model, Genie 4, in a research preview. The latest model generates photorealistic and interactive worlds, allowing users to type prompts mid-simulation to alter the scene in real time. Alongside visual updates, the system can generate synchronized audio that matches the simulated environment, including music, ambient noises, and interactive dialogue.
The release builds upon previous iterations like the earlier Genie 3 model. Google announced its Genie 4 world model for generating interactive experiences using AI, highlighting updates that include enhanced realism, real-time natural language updates, audio generation, and the use of Gemini 3.8 Flash to assist with simulations.
New Capabilities and Technical Updates
Genie 4 introduces significant improvements over its predecessor by doubling the duration limit for maintaining consistent scenes. While earlier versions supported shorter runtimes, Genie 4 can maintain a consistent scene for up to two minutes. The model also offers better physics and more humanlike characters.
Furthermore, the system allows for free-form edits, enabling users to change weather conditions, adjust camera angles, or introduce new people and objects mid-simulation. Google notes that Gemini acts closely with the system to plan steps and check frames while visuals are created. In a research experiment utilizing Gemini Robotics, a Gemini 4 Argon model broke down tasks into steps that Genie 4 simulated in real time.
Commercial Pitch and Potential Use Cases
Google is pitching the software as a helpful tool across several industries, including pre-visualization for filmmakers, game idea prototyping, and creating training scenarios for robots. Engineers and researchers note that the real-time aspect proves especially valuable during ideation and storyboarding phases where maintaining creative flow is crucial.
Additional speculative use cases span gaming and interactive learning. For instance, an educational simulation might allow users to step inside a historical painting and converse directly with the subjects portrayed within the scene.
Industry Context and Limitations
The arrival of Genie 4 follows other developments in the competitive world model sector, arriving a couple of months after AI video startup Runway released its second world model, GWM Worlds 2. Meanwhile, broader industry consolidation continues as various companies acquire simulation and world-building startups for autonomous driving and virtual testing.
Despite these advancements, Google's team acknowledges that the technology remains in early stages. As simulations run longer, details and object permanence can shift over time, and developers are still working to ensure audio accuracy and text legibility inside the generated worlds.
Availability and Access
Genie 4 is currently available as a research preview restricted to trusted testers. Google has stated that the model will soon roll out to Google AI Ultra subscribers through Project Genie, as well as to early access partners of the Gemini API and Google AI Studio.
Users interested in trying out the technology can sign up via a waitlist to apply for early access to Project Genie.
Sources
- The VergeGoogle’s next Genie AI model generates more realistic experiences
- TechCrunchGoogle’s Genie 4 world model adds audio, more realistic worlds, and real-time steering
Continue chronologically
Related entity coverage
Related stories
Google DeepMind Unveils Genie 4 World Model
Google DeepMind has introduced Genie 4, a new world model featuring real-time audio generation, improved physics, and more realistic human characters.
Google DeepMind Unveils Genie 4 World Simulation Model
Google DeepMind introduces Genie 4, a new real-time interactive world simulation model featuring native audio, advanced physics, and Gemini integration.
Google Introduces Gemini 3.8 Live with New Live Avatar
Google DeepMind launches Gemini 3.8 Live with Live Avatar, bringing near real-time visual presence and multilingual support to enterprise AI agents.
Google Opens Gemini and Search to More Apps With Connected Platform
Google debuts a new developer platform allowing businesses to connect their third-party apps directly to Gemini and Google Search.
