The Matrix: Infinite-Horizon World Generation with Real-Time Moving Control

Yu Liu (Alibaba Group) · Hongyang Zhang (School of Computer Science, University of Waterloo) · Ruihang Chu (Wan, Alibaba Group) · Han Zhang (Shanghai Jiao Tong University) · Zhiheng Liu (The University of Hong Kong) · Ruili Feng (University of Waterloo) · Zhilei Shu (University of Science and Technology of China) · Zhantao Yang (Alibaba Group) · Longxiang Tang (Tsinghua University) · Zhicai Wang (University of Science and Technology of China) · Andy Zheng (University of Waterloo) · Jie Xiao (University of Science and Technology of China) · Yukun Huang (University of Science and Technology of China)
continuous movement datadiverse terrainsfirst-person perspectivefoundational simulatorhigh-fidelity video streamslimited data scenariosreal-time controlrobust world modelssimulation bridgingsupervised datathird-person perspectiveunsupervised footagevirtual game environmentszero-shot generalization

We present The Matrix, a foundational realistic world simulator capable of generating infinitely long 720p high-fidelity real-scene video streams with real-time, responsive control in both first- and third-person perspectives. Trained on limited supervised data from video games like Forza Horizon 5 and Cyberpunk 2077, complemented by large-scale unsupervised footage from real-world settings like Tokyo streets, The Matrix allows users to traverse diverse terrains—deserts, grasslands, water bodies, and urban landscapes—in continuous, uncut hour-long sequences. With speeds of up to 16 FPS, the system supports real-time interactivity and demonstrates zero-shot generalization, translating virtual game environments to real-world contexts where collecting continuous movement data is often infeasible. For example, The Matrix can simulate a BMW X3 driving through an office setting—an environment present in neither gaming data nor real-world sources. This approach showcases the potential of game data to advance robust world models, bridging the gap between simulations and real-world applications in scenarios with limited data.