Based in New York City, JoJoWorld is a spatial 3D data platform that supports the training of embodied AI systems, text-to-3D models, and large world models. The project develops decentralized infrastructure to acquire, curate, and annotate high-quality 3D data at scale.

JoJoWorld addresses the 3D data bottleneck in advanced AI applications, particularly those requiring spatial awareness and physical-world interaction. To solve this, the platform incentivizes global contributors to upload raw 3D assets and metadata via token-based rewards. These assets are enhanced through JoJoWorld’s proprietary Gaussian Splatting pipelines and AI-assisted auto-labeling tools.

The ecosystem serves both individual 3D creators and institutional AI developers. Contributors earn tokens by uploading 3D scans, photogrammetry data, or other spatial inputs. Enterprises and AI firms can access JoJoWorld’s data via commercial agreements or ecosystem integrations. Use cases include robot training, simulation environments, generative models, and virtual worlds.

As of September 2025, JoJoWorld is undergoing commercialization and has already generated revenue through partnerships with AI and robotics companies. Clients and collaborators include teams developing embodied AI agents, humanoid robots, and vision-language-3D foundation models.

JoJoWorld’s team includes a tenured NYU professor specializing in AI as co-founder, CTO Wayne Li—a Silicon Valley entrepreneur with two prior exits—and a core team with backgrounds from Stanford, Carnegie Mellon, Harvard, the World Bank, Goldman Sachs, Citigroup, Meta, YouTube, and OKX Global.

The platform has been selected into multiple accelerator programs, including Binance’s MVB9, Solana’s SpringX, and Microsoft for Startups. Its infrastructure is built on a DeAI model, enabling community-driven data contribution, validation, and participation in the data economy.