Wan 2.2
Alibaba’s open-weight video generation models (text, image and speech to video).
- Star 数
- 17.9K
- Fork 数
- 2.3K
- 未关闭 Issue
- 318
- 最近提交
- 2026-09-21
数据来自 GitHub API · 更新于 2026-10-15
Wan 2.2 是什么？
Wan 2.2 is Alibaba’s open video model family under Apache 2.0. It includes large Mixture-of-Experts text-to-video and image-to-video models and a compact 5B text+image-to-video model that runs on a single consumer GPU, plus ComfyUI and Diffusers integrations.
主要功能
- Text-to-video and image-to-video at 480p/720p
- TI2V-5B model fits a 24 GB consumer GPU
- Mixture-of-Experts A14B models for higher quality
- ComfyUI and Diffusers support
系统要求
- 平台
- Linux · Windows
- 显卡
- NVIDIA
- 显存
- 24 GB for TI2V-5B at 720p; 80 GB-class GPUs (or offloading) for A14B models
- Python
- 3.10+
如何安装 Wan 2.2
根据项目 README 整理。开源项目更新很快，请以 GitHub 页面上的最新说明为准。
Reference code
- 1Clone and install (PyTorch ≥ 2.4).
git clone https://github.com/Wan-Video/Wan2.2.git cd Wan2.2 pip install -r requirements.txt - 2Download a model from Hugging Face.
huggingface-cli download Wan-AI/Wan2.2-TI2V-5B --local-dir ./Wan2.2-TI2V-5B - 3Generate a clip.
python generate.py --task ti2v-5B --size 1280*704 --ckpt_dir ./Wan2.2-TI2V-5B --offload_model True --convert_model_dtype --t5_cpu --prompt "A cat surfing at sunset"
ComfyUI
- 1Update ComfyUI and open the Wan 2.2 workflow from the built-in templates; it lists the model files to download.
优势
- Best open-weight video quality class
- Apache 2.0 license
- Runs locally
不足
- Large models need data-center GPUs
- Slow on consumer hardware
Wan 2.2 替代品对比
| 软件 | Star 数 | 许可证 | 平台 | 显存 |
|---|---|---|---|---|
| Wan 2.2 Alibaba’s open-weight video generation models (text, image and speech to video). | 17.9K | Apache-2.0 | 24 GB for TI2V-5B at 720p; 80 GB-class GPUs (or offloading) for A14B models | |
| HunyuanVideo Tencent’s 13B-parameter open video foundation model. | 12.6K | Tencent Hunyuan Community License | 45 GB minimum, 60–80 GB recommended for 720p (per README); less with community quantizations | |
| LTX-Video Fast, real-time-capable open video generation from Lightricks. | 11.1K | Apache-2.0 | 12 GB+ for smaller variants; more for 13B models | |
| CogVideoX Open text-to-video and image-to-video models from Zhipu AI / Tsinghua. | 13.1K | Apache-2.0 | ~12–24 GB depending on model and optimizations (see the README table) | |
| FramePack Long image-to-video generation on laptop GPUs with as little as 6 GB VRAM. | 17.3K | Apache-2.0 | 6 GB minimum | |
| AnimateDiff Animate your Stable Diffusion checkpoints with plug-in motion modules. | 12.3K | Apache-2.0 | 8–12 GB |
相关模型
常见问题
Wan 2.2 免费吗？
是的——Wan 2.2 是开源软件，采用 Apache-2.0 许可证。商用前请确认软件及模型权重的许可条款。
Wan 2.2 支持 Windows、Mac 和 Linux 吗？
Wan 2.2 支持：Linux, Windows。
运行 Wan 2.2 需要什么配置？
NVIDIA. VRAM: 24 GB for TI2V-5B at 720p; 80 GB-class GPUs (or offloading) for A14B models
如何安装 Wan 2.2？
1. Clone and install (PyTorch ≥ 2.4). 2. Download a model from Hugging Face. 3. Generate a clip.
Wan 2.2 有哪些好的替代品？
常见替代品包括 HunyuanVideo, LTX-Video, CogVideoX, FramePack，详见本页对比表。
可以不安装、在线使用 Wan 2.2 吗？
如果没有大显存显卡，AiLoft 可在云端运行 Veo 3.1 和可灵 2.5，在浏览器中生成并对比视频。
没有 24GB 显卡？在线生成 AI 视频
在云端运行 Veo 3.1 和可灵 2.5，并排对比视频，按条付费。