---
license: apache-2.0
datasets:
- nkp37/OpenVid-1M
language:
- en
base_model:
- Wan-AI/Wan2.1-T2V-1.3B-Diffusers
pipeline_tag: text-to-video
tags:
- iclr
- wan
- qat
- videagen
---

<div align="center" style="font-family: charter;">
<h1> QVGen:<br> Pushing the Limit of Quantized Video Generative Models</h1>

[![License](https://img.shields.io/badge/License-Apache_2.0-blue.svg)](https://opensource.org/licenses/Apache-2.0)&nbsp;
[![arXiv](https://img.shields.io/badge/QVGen-2505.11497-b31b1b)](https://arxiv.org/pdf/2505.11497)&nbsp;
[![Hugging Face](https://img.shields.io/badge/🤗%20Hugging%20Face-Models-yellow)](https://huggingface.co/collections/Harahan/qvgen)&nbsp;

**[ [Conference Paper](https://arxiv.org/abs/2505.11497) | [Models](https://huggingface.co/collections/Harahan/qvgen) | [Dataset](https://huggingface.co/datasets/nkp37/OpenVid-1M) | [Code](https://github.com/ModelTC/QVGen) ]**

[Yushi Huang](https://Harahan.github.io/), [Ruihao Gong📧](https://xhplus.github.io/), [Jing Liu](https://jing-liu.com/), [Yifu Ding](https://yifu-ding.github.io/), [Chengtao Lv](https://scholar.google.com/citations?user=r8vseSUAAAAJ&hl=en), [Haotong Qin](https://htqin.github.io/), [Jun Zhang📧](https://eejzhang.people.ust.hk/)

(📧 denotes corresponding author.)

</div>

## 📖 Overview

[QVGen](https://arxiv.org/abs/2505.11497) is *the first* to reach full-precision comparable quality under 4-bit settings and it significantly outperforms existing methods. For instance, our 3-bit CogVideoX-2B improves Dynamic Degree by +25.28 and Scene Consistency by +8.43 on VBench.

## ⚙️ Usage

See our official [code base](https://github.com/ModelTC/QVGen).

## ✨ Model Zoo

| Model | #Bit |
| --- | --- |
| [Wan 1.3B](https://huggingface.co/Harahan/QVGen-Wan-1_3B-W4A4) | W4A4 |
| [CogVideoX-2B](https://huggingface.co/Harahan/QVGen-CogVideoX-2B-W4A4) | W4A4 |

## ✏️ Citation

If you find QVGen useful, please cite our paper:

```
@inproceedings{huang2026qvgenpushinglimitquantized,
  title={QVGen: Pushing the Limit of Quantized Video Generative Models}, 
  author={Yushi Huang and Ruihao Gong and Jing Liu and Yifu Ding and Chengtao Lv and Haotong Qin and Jun Zhang},
  booktitle={International Conference on Learning Representations},
  year={2026},
  url={https://arxiv.org/abs/2505.11497}, 
}
```