AI

Google DeepMind Announces Genie 4 World-Model Research Preview

Tuesday, October 27, 2026Read Original

Details

  • Google DeepMind announced Genie 4, a new world model for real-time interactive simulation.
  • The system generates dynamic environments with visual detail, environmental audio, music, and interactive dialogue.
  • Gemini Omni supplies realistic appearances and behavior, while other Gemini models help maintain scene coherence and continuity.
  • Users can steer simulations with natural-language prompts, including changing weather, moving the camera, or adding people and objects.
  • Genie 4 is designed to model physics such as aerodynamics and collisions, while producing more natural human movement and reactions.
  • The rollout began as a research preview for trusted testers, with broader access planned through Project Genie, the Gemini API, and Google AI Studio.
  • Google says the system is intended to support interactive experiences rather than pre-rendered environments, generating scenes as users explore and modify them.

Impact

Genie 4 extends Google’s Gemini strategy from generating media to simulating persistent, steerable environments. Native audio, live scene editing, and improved physics could broaden world models’ uses in entertainment, training, robotics research, and prototyping, while the limited preview signals that reliability, safety, latency, and infrastructure costs remain unresolved. Planned access through Project Genie and developer tools could also give Google a distribution advantage as rivals pursue comparable interactive video and simulation systems.

Rift Dispatch
Google DeepMind Announces Genie 4 World-Model Research Preview | RiftLab