3 min read
Google has introduced Project Genie, a new experimental artificial intelligence system capable of generating interactive three-dimensional virtual worlds from text and image prompts, marking a notable expansion of generative AI beyond static content like text, images, and video. The tool is being rolled out to U.S. users subscribed to Google’s AI Ultra plan, positioning it as an early-access product rather than a mass-market launch.
According to the search engine tech giant:
In August, we previewed Genie 3, a general-purpose world model capable of generating diverse, interactive environments. Even in this early form, trusted testers were able to create an impressive range of fascinating worlds and experiences, and uncovered entirely new ways to use it. The next step is to broaden access through a dedicated, interactive prototype focused on immersive world creation.
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.
Project Genie is built on Genie 3, a world-model developed by Google DeepMind that can simulate environments dynamically as users explore them. Unlike traditional game engines that rely on pre-built assets, Genie 3 generates terrain, structures, and scenery on the fly, responding in real time to user movement and camera input. Current sessions run at approximately 720p resolution and 24 frames per second, with individual world interactions limited to about one minute.
According to Google:
Project Genie is a prototype web app powered by Genie 3, Nano Banana Pro and Gemini, which allows users to experiment with the immersive experiences of our world model firsthand. The experience is centred on three core capabilities:
1. World sketching
2. World exploration
3. World remixing
To create a world, users provide a short description of the environment and optionally define an avatar. Project Genie first produces a rough visual concept using Google’s image models, then converts that input into a navigable 3D space.
Users can walk, fly, or move freely through these environments, remix existing worlds, and export short video clips of their sessions. Google describes the experience as a “world waiting to be explored,” emphasizing interaction rather than passive viewing.
Here is where Project Genie has an edge over Unreal and Unity gaming world-building systems, which need:
- Asset creation (models, textures, lighting)
- Scene assembly
- Physics tuning
- Scripting and debugging
- Long iteration cycles
Project Genie collapses all of this into one prompt. On the other hand, despite all the glory, Project Genie remains clearly experimental. As Google puts it:
Since Genie 3 is an early research model, there are a few known areas for improvement:
- Generated worlds might not look completely true-to-life or always adhere closely to prompts or images, or real-world physics
- Characters can sometimes be less controllable, or experience higher latency in control
- Limitations in generations to 60 seconds
The tool is currently available in the U.S. for users who have a Google AI Ultra subscription. This plan, which costs $250 (i.e., PKR ~70K) a month, comes packed with features that are not part of the standard Google accounts. Subscribers enjoy higher limits on AI usage, a whopping 30TB of cloud storage, and access to a faster version of the company’s Antigravity agentic coding tool.
