We introduce WorldGen, a system that enables the automatic creation of large-scale, interactive 3D worlds directly from text prompts. Our approach transforms natural language descriptions into traversable, fully textured environments that can be immediately explored or edited within standard game engines. By combining LLM-driven scene layout reasoning, procedural generation, diffusion-based 3D generation, and object-aware scene decomposition, WorldGen bridges the gap between creative intent and functional virtual spaces, allowing creators to design coherent, navigable worlds without manual modeling or specialized 3D expertise. The system is fully modular and supports fine-grained control over layout, scale, and style, producing worlds that are geometrically consistent, visually rich, and efficient to render in real time. This work represents a step towards accessible, generative world-building at scale, advancing the frontier of 3D generative AI for applications in gaming, simulation, and immersive social environments.
@article{arxiv.2511.16825,
title = {WorldGen: From Text to Traversable and Interactive 3D Worlds},
author = {Dilin Wang and Hyunyoung Jung and Tom Monnier and Kihyuk Sohn and Chuhang Zou and Xiaoyu Xiang and Yu-Ying Yeh and Di Liu and Zixuan Huang and Thu Nguyen-Phuoc and Yuchen Fan and Sergiu Oprea and Ziyan Wang and Roman Shapovalov and Nikolaos Sarafianos and Thibault Groueix and Antoine Toisoul and Prithviraj Dhar and Xiao Chu and Minghao Chen and Geon Yeong Park and Mahima Gupta and Yassir Azziz and Rakesh Ranjan and Andrea Vedaldi},
journal= {arXiv preprint arXiv:2511.16825},
year = {2025}
}