Google has officially launched Project Genie, a web-based prototype that allows users to create and explore 3D worlds generated by artificial intelligence from just text commands or images. This technology opens up new ways of interacting with AI, from simply creating static images to exploring a living and dynamic virtual world.
Project Genie is an user interface version of Genie 3, Google's latest AI model designed to generate three-dimensional environments. The project also utilizes the Gemini model as well as Google's AI image-making technology Nano Banana.
According to Google, the process of creating the world in Project Genie starts with the World Sketching feature. At this stage, users can describe the desired world using natural language or upload images as visual references.
Google explained that the World Sketching feature has been integrated directly with Nano Banana, so users can first compose a visual concept before the AI unites it into a complete 3D world through Project Genie.
After the world is formed, users can enter the World Exploration stage. Here, AI will continue to generate environments in front of users in real time as they move, resembling the concept of an open world game without a fixed map.
In addition, there is also a World Remixing feature that allows users to take someone else's artificial world and develop it again with new prompts. In other words, one world can evolve into another world just from the user's additional ideas.
Introducing Project Genie: An experimental research prototype powered by Genie 3, our world model, that lets you prompt an interactive world into existence — and then step inside 🌎 pic.twitter.com/Zn84k0iJ48
— Google (@Google) January 29, 2026
Although it sounds revolutionary, Project Genie's access is still very limited. Google only opens this service to AI Ultra customers in the United States. The AI Ultra package is Google's most expensive AI subscription with a cost of up to 250 US dollars per month, making it almost unaffordable for general users.
Even for AI Ultra customers, Project Genie still has a number of technical limitations. One exploration session is limited to a maximum of 60 seconds. The generated world also only appears in 720p resolution, with an interaction speed of around 20 to 24 frames per second, which means it's not completely smooth.
Even so, Project Genie is still considered a major leap forward in generative AI development. Instead of just creating images or videos, this technology allows AI to build an interactive world that can be explored, modified, and continuously developed.
If this technology continues to be developed and is widely available, Project Genie has the potential to change the way games, simulations, virtual worlds, and AI-based educational experiences are created. For now, the future of the AI-based artificial world is still the exclusive right of "upper class" users.
The English, Chinese, Japanese, Arabic, and French versions are automatically generated by the AI. So there may still be inaccuracies in translating, please always see Indonesian as our main language. (system supported by DigitalSiber.id)
