---
task_categories:
- video-text-to-text
language:
- en
---

This dataset is for evaluating world modeling capabilities of video generation models, as described in the paper [WorldModelBench: Judging Video Generation Models As World Models](https://hf.co/papers/2502.20694).

*   **Project Page:** [https://worldmodelbench-team.github.io](https://worldmodelbench-team.github.io/)