The Year in AI Papers
The year/Independent research

Paper 2512.14614

WorldPlay: Towards Long-Term Geometric Consistency for Real-Time Interactive World Modeling

Wenqiang Sun, Haiyu Zhang, Haoyuan Wang, Junta Wu, Zehan Wang, Zhenwei Wang, Yunhong Wang, Jun Zhang, Tengfei Wang, Chunchao Guo

Published
Dec 2025
Research lab
Independent
Citations
97
GitHub
1.6K stars

01 In brief

Summary

WorldPlay is a real-time interactive world model that generates streaming 720p video at 24 FPS while maintaining long-term geometric consistency.

It addresses the trade-off between speed and memory in existing methods.

The model uses three key components: Dual Action Representation combining discrete keys and continuous camera poses for robust control; Reconstituted Context Memory that dynamically rebuilds context from past frames and uses temporal reframing to keep long-past geometrically important frames accessible; and Context Forcing, a distillation method that aligns memory context between teacher and student to enable real-time speeds while preventing error drift.

WorldPlay is trained on a dataset of 320K real and synthetic videos and demonstrates superior performance over baselines in quantitative metrics (PSNR, SSIM, LPIPS, R_dist, T_dist) and qualitative comparisons, especially in long-term settings.

It generalizes across diverse scenes including first- and third-person real and stylized worlds, and supports applications like 3D reconstruction and text-promptable events.

The model achieves real-time interactivity through optimizations including mixed parallelism, streaming deployment, progressive decoding, quantization, and KV-cache mechanisms.

02 From the paper

Abstract

This paper presents WorldPlay, a streaming video diffusion model that enables real-time, interactive world modeling with long-term geometric consistency, resolving the trade-off between speed and memory that limits current methods. WorldPlay draws power from three key ingredients. 1) We use a Dual Action Representation to enable robust action control in response to the user's keyboard and mouse inputs. 2) To enforce long-term consistency, our Reconstituted Context Memory dynamically rebuilds context from past frames and uses temporal reframing to keep geometrically important but long-past frames accessible, effectively alleviating memory attenuation. 3) We also propose Context Forcing, a novel distillation method designed for memory-aware model. Aligning memory context between the teacher and student preserves the student's capacity to use long-range information, enabling real-time speeds while preventing error drift. Taken together, WorldPlay generates long-horizon streaming 720p video at 24 FPS with superior consistency, comparing favorably with existing techniques and showing strong generalization across diverse scenes. Project page and online demo can be found: https://3d-models.hunyuan.tencent.com/world/ and https://3d.hunyuan.tencent.com/sceneTo3D.