Google has made Project Genie available to Premium Gemini subscribers in the US. This lets users create, explore, and remix their own dynamic worlds. This project is an experimental research prototype that uses the Genie 3 AI model. Its goal is to enable the creation of fully interactive virtual environments using natural-language input. Users can create situations that last for up to 60 seconds. Project Genie comes after the release of Genie 3 in August 2025 and adds features from Nano Banana and Gemini 3.
“Starting today, we’re rolling out access to Project Genie for Google AI Ultra subscribers in the U.S. (18+). This experimental research prototype lets users create, explore, and remix their own interactive worlds,” the company said via blog post.
In This Article
Capabilities
Project Genie is a new prototype web app that uses Genie 3, Nano Banana Pro, and Gemini to let users interact directly with immersive experiences based on a world model.
Users can make and change surroundings with the World sketching tool by typing in text and using images that are either created or uploaded. This involves making your own characters and choosing how to explore, such as walking, flying, or driving. When you connect with Nano Banana Pro, this process improves because you can see and change things in real time to enhance both the visual landscape and the user’s perspective, whether in first- or third-person.
The application makes the world a dynamic place to explore, with paths that change in real time as users interact with it. Users can change the angles of the cameras to make their discovery more fun.
Users can remix existing worlds, enabling them to generate new ideas by modifying existing prompts. They can also browse curated galleries for inspiration and download videos of their creations and explorations once finished.
Also Read: Union Budget 2026 Major Tech Announcements
Limitations
The system’s current constraints include a limited action space, which means that agents can only do a small number of things, even when there are many world events that could urge them to do so. The interaction and simulation of several autonomous agents constitute a continuing research challenge. Also, Genie 3 can’t fully imitate real-world locations, and clear text rendering only works in the world description context. The model also only works for short-duration interactions, which usually last a few minutes.
Pricing and Availability
In the United States, Google AI Ultra customers can use Project Genie, which is an interactive feature that is only available to anyone over the age of 18. The Ultra subscription package costs $249.99 a month in the U.S., which is about Rs 22,980. The same subscription in India costs Rs 24,500 a month, but it’s important to keep in mind that Project Genie is not currently available in India.
