---
license: apache-2.0
---
## X-Omni-Zh (support Chinese text rendering)

<p align="left">
  <a href="https://x-omni-team.github.io">🏠 Project Page</a> |
  <a href="https://arxiv.org/pdf/2507.22058">📄 Paper</a> |
  <a href="https://github.com/X-Omni-Team/X-Omni">💻​ Code</a> |
  <a href="https://huggingface.co/collections/X-Omni/x-omni-spaces-6888c64f38446f1efc402de7">🚀 HuggingFace Space</a> 
</p>

## 🌟 Highlights

- **Unified Modeling Approach**: A discrete autoregressive model handling image and language modalities.
- **Superior Instruction Following**: Exceptional capability to follow complex instructions.
- **Superior Text Rendering**: Accurately render text in Chinese.
- **Arbitrary resolutions**: Produces aesthetically pleasing images at arbitrary resolutions.

<p align="left">
    <img src="assets/fig2-1.png" alt="" width="600" />
    <img src="assets/fig5-1.png" alt="" width="600" />
</p>

## 📖 Citation

If you find this project helpful for your research or use it in your own work, please cite our paper:
```bibtex
@article{geng2025xomni,
      author       = {Zigang Geng, Yibing Wang, Yeyao Ma, Chen Li, Yongming Rao, Shuyang Gu, Zhao Zhong, Qinglin Lu, Han Hu, Xiaosong Zhang, Linus, Di Wang and Jie Jiang},
      title        = {X-Omni: Reinforcement Learning Makes Discrete Autoregressive Image Generative Models Great Again},
      journal      = {CoRR},
      volume       = {abs/2507.22058},
      year         = {2025},
}
```