About - Open-Sora

Our Mission

Open-Sora is an open-source project dedicated to democratizing efficient video production. We believe that high-quality video generation should be accessible to everyone, not just large corporations with massive resources.

Our goal is to provide state-of-the-art video generation capabilities through an open-source model that can be used, modified, and improved by the community.

What is Open-Sora?

Open-Sora is a powerful text-to-video and image-to-video generation model built on diffusion transformers. It enables users to generate high-quality videos from simple text prompts or static images.

Open-Sora 2.0 represents a significant advancement, narrowing the performance gap with commercial models while remaining cost-effective and open-source.

Key Achievements

Performance: Reduced gap with OpenAI's Sora from 4.52% to 0.69% on VBench
Cost-Effective: Training completed in $200k, making it accessible
Community: Over 28.4k GitHub stars and 57+ contributors
Open Source: Apache 2.0 license for maximum accessibility
Flexibility: Support for multiple resolutions and aspect ratios

Technology Stack

Open-Sora is built on cutting-edge technologies:

ColossalAI: Powerful large model parallel acceleration system
DiT: Scalable Diffusion Models with Transformers
OpenDiT: Acceleration strategies for training
PixArt: DiT-based text-to-image model foundation
Flux: Powerful text-to-image generation model

Acknowledgments

Open-Sora builds upon the work of many excellent open-source projects and datasets:

ColossalAI for parallel acceleration
DiT and OpenDiT for diffusion transformer architecture
PixArt and Flux for image generation foundations
Latte for efficient DiT training approaches
StabilityAI VAE and DC-AE for image compression
CLIP and T5 for text-image embeddings
LLaVA and PLLaVA for captioning capabilities
MiraData for large-scale video datasets

Contributing

We welcome contributions from the community! Whether you're fixing bugs, adding features, improving documentation, or sharing examples, your contributions help make Open-Sora better for everyone.

To get started:

Fork the repository on GitHub
Create a branch for your changes
Make your improvements
Submit a pull request

For more details, please refer to our Contribution Guidelines.

Citation

If you use Open-Sora in your research, please cite:

@article{opensora,
  title={Open-sora: Democratizing efficient video production for all},
  author={Zheng, Zangwei and Peng, Xiangyu and Yang, Tianji and Shen, Chenhui and Li, Shenggui and Liu, Hongxin and Zhou, Yukun and Li, Tianyi and You, Yang},
  journal={arXiv preprint arXiv:2412.20404},
  year={2024}
}

@article{opensora2,
    title={Open-Sora 2.0: Training a Commercial-Level Video Generation Model in $200k}, 
    author={Xiangyu Peng and Zangwei Zheng and Chenhui Shen and Tom Young and Xinying Guo and Binluo Wang and Hang Xu and Hongxin Liu and Mingyan Jiang and Wenjun Li and Yuhui Wang and Anbang Ye and Gang Ren and Qianran Ma and Wanying Liang and Xiang Lian and Xiwen Wu and Yuting Zhong and Zhuangyan Li and Chaoyu Gong and Guojun Lei and Leijun Cheng and Limin Zhang and Minghao Li and Ruijie Zhang and Silan Hu and Shijie Huang and Xiaokang Wang and Yuanheng Zhao and Yuqi Wang and Ziang Wei and Yang You},
    year={2025},
    journal={arXiv preprint arXiv:2503.09642},
}

License

Open-Sora is licensed under the Apache 2.0 License. This means you are free to use, modify, and distribute the software, even for commercial purposes.

About Open-Sora