Description
Project Genie is an AI world generation platform that lets creators build explorable 3D worlds from text prompts and image references. Users start with a prompt, optionally add image references, choose a world model, and generate scenes inside a web-based studio.
The platform supports a range of environments, including fantasy scenes, sci-fi spaces, landscapes, architectural concepts, and early game-world ideas. It offers multiple generation modes so creators can trade off speed and quality, using faster models for rough drafts and higher-quality models for polished outputs.
Generation is managed through credits, with model costs shown up front so users can plan drafts, revisions, and final runs. After generating, worlds can be explored in the studio, kept private for iteration, or published publicly. Project Genie is designed for a prompt-first workflow that requires no technical skills, and it is an independent product not affiliated with Google or Google DeepMind.
Project Genie's Core Features
Text-to-world generation from prompts
Image-guided world creation with reference images
Multiple fast and high-quality world models
Studio for exploring generated worlds
Credit-aware workflow with costs shown up front
Private iteration or public publishing
Community discovery of published worlds
Commercial use license on paid plans
How to use Project Genie?
Describe your world: Enter a text prompt on the Create page describing the scene you want.
Add references: Optionally include one or more image references for stronger visual guidance.
Choose a model: Select a fast draft model or a higher-quality model based on your goal and credits.
Explore and interact: Open the result in the studio to review and continue exploring.
Publish: Keep the world private for iteration or publish it publicly to share.
Project Genie's Use Cases
- Game development
- Education and training
- Film and entertainment
- Architecture and design








