Google has launched Project Genie, a new web-based platform that lets people build AI-powered 3D interactive worlds using text prompts and images. The service is based on DeepMind’s Genie 3 model, first introduced last year as a research tool for training AI agents.
Project Genie is available through the Google AI Ultra subscription, which costs USD 250 per month (PHP 13,900 in the Philippines via Google One). For now, access is limited to users in the United States who are 18 years old and above.
The platform offers three modes. World Sketching allows users to describe the world, the characters, perspectives, and exploration styles. Google’s Nano Banana Pro model then generates a sketch before building the environment. Exploration lets users move through the created world. Remixing enables users to modify or reuse worlds made by others.
Take note that Project Genie is not a full game engine. It does not include gaming mechanics like scoring or combat. Instead, it focuses on world modeling and creative experimentation. Each generation is capped at 60 seconds of exploration, with a maximum resolution of 720p and a frame rate of 24 fps.
Also Read: Chrome adds Gemini 3 AI features, Auto Browse included
Project Genie offers users a chance to experiment with interactive environments that respond to movement and prompts.
What do you think of Google’s latest service? Is this something you’d like to try? Stay tuned for updates.
