Google DeepMind announced the limited rollout of Project Genie, an experimental research prototype that enables users to create, explore, and remix immersive interactive worlds using generative AI. Initially available to Google AI Ultra subscribers in the United States (18+), Project Genie expands access to Google’s world model technology and provides a first-hand experience of next-generation environment generation tools built by DeepMind and Google Labs.
Project Genie is driven by Genie 3, a general-purpose world model that has the ability to create dynamic and navigable worlds in real-time as the users interact with them. Using text prompts and images, users are able to draw worlds, navigate through them, and even remix other people’s worlds to create new ones.
Project Genie combines several advanced technologies across the Google AI stack, including Genie 3, Nano Banana Pro, and Gemini, to deliver an interactive world-building experience in a prototype web application. The experience centers on three core capabilities:
World sketching, which allows users to create environments based on text and image prompts, define avatars and perspectives, and fine-tune world details prior to exploration.
Also Read: IBM Accelerates a Vibrant AI Ecosystem to Drive Growth and Partner Success
World exploration, enabling users to traverse environments that generate paths ahead in real time based on user actions.
World remixing, letting users build on top of existing worlds or discover inspiration in galleries and randomized experiences.
The prototype also allows users to download video captures of their worlds and experiences for sharing or review. While the current experience reflects early research conditions, including known limitations in realism and character control, Google DeepMind intends to gather user feedback and iterate on the technology over time.
Google emphasized that Project Genie remains an experimental research effort developed within Google Labs, and that generative AI technologies like world models are still under active investigation. As part of the company’s commitment to building AI responsibly and safely to benefit humanity, ongoing improvements will address technical limitations and user interaction challenges.
Project Genie is now available to eligible subscribers. It will expand to more users and regions over time. Google is excited to see the variety of worlds that the research community will create in future development efforts.