WorldVLA is an autoregressive action world model that unifies action and image understanding and generation. WorldVLA intergrates Vision-Language-Action (VLA) model (action model) and world model in one single framework.
Action Model generates actions given the text instruction and image observations.
World Model generates the next frame given the current frame and action control.
If you find the project helpful for your research, please consider citing our paper:
1@article{cen2025worldvla,
2 title={WorldVLA: Towards Autoregressive Action World Model},
3 author={Cen, Jun and Yu, Chaohui and Yuan, Hangjie and Jiang, Yuming and Huang, Siteng and Guo, Jiayan and Li, Xin and Song, Yibing and Luo, Hao and Wang, Fan and others},
4 journal={arXiv preprint arXiv:2506.21539},
5 year={2025}
6}
This project builds upon
Lumina-mGPT,
Chemeleon, and
OpenVLA. We thank these teams for their open-source contributions.
