{"name":"Wan2.2: Open and Advanced Large-Scale Video Generative Models","description":"Wan2.2 is an open-source and advanced suite of large-scale video generative models, introducing innovations like a Mixture-of-Experts (MoE) architecture for enhanced capacity and cinematic-level aesthetics. It offers efficient high-definition video generation capabilities, including text-to-video, image-to-video, speech-to-video, and character animation. This powerful framework is designed for both industrial and academic applications, pushing the boundaries of AI-driven video creation.","github":"https://github.com/Wan-Video/Wan2.2","url":"https://osrepos.com/repo/wan-video-wan2.2","source":"osrepos.com","sourceDescription":"This repository profile is provided by osrepos.com, an open source repository discovery platform.","repositoryProfile":"https://osrepos.com/repo/wan-video-wan2.2","generatedFor":"open source discovery and AI-assisted research","markdown":"https://osrepos.com/repo/wan-video-wan2.2.md","json":"https://osrepos.com/repo/wan-video-wan2.2.json","topics":["aigc","video-generation","Python","generative AI","deep learning","MoE","video models"],"keywords":["aigc","video-generation","Python","generative AI","deep learning","MoE","video models"],"stars":null,"summary":"Wan2.2 is an open-source and advanced suite of large-scale video generative models, introducing innovations like a Mixture-of-Experts (MoE) architecture for enhanced capacity and cinematic-level aesthetics. It offers efficient high-definition video generation capabilities, including text-to-video, image-to-video, speech-to-video, and character animation. This powerful framework is designed for both industrial and academic applications, pushing the boundaries of AI-driven video creation.","content":"## Introduction\n\nWan2.2 is an open and advanced suite of large-scale video generative models, pushing the boundaries of AI-driven video creation. This major upgrade builds upon its predecessor, Wan2.1, by incorporating several key innovations designed to enhance generation quality, model capability, and computational efficiency. It introduces a Mixture-of-Experts (MoE) architecture, achieves cinematic-level aesthetics through meticulously curated data, and significantly improves complex motion generation with a larger training dataset. Furthermore, Wan2.2 offers an efficient high-definition hybrid Text-Image-to-Video (TI2V) model, capable of generating 720P videos at 24fps on consumer-grade GPUs.\n\n## Installation\n\nTo get started with Wan2.2, follow these simple steps:\n\nFirst, clone the repository:\n\nsh\ngit clone https://github.com/Wan-Video/Wan2.2.git\ncd Wan2.2\n\n\nNext, install the required dependencies. Ensure you have `torch >= 2.4.0`. If `flash_attn` installation fails, try installing other packages first and then `flash_attn`.\n\nsh\npip install -r requirements.txt\n# If you want to use CosyVoice for Speech-to-Video Generation, install additional requirements:\npip install -r requirements_s2v.txt\n\n\n## Examples\n\nWan2.2 supports various video generation tasks, including Text-to-Video (T2V), Image-to-Video (I2V), Text-Image-to-Video (TI2V), Speech-to-Video (S2V), and character animation with Wan-Animate.\n\nYou can download the models using `huggingface-cli` or `modelscope-cli`. For instance, to download the T2V-A14B model:\n\nsh\npip install \"huggingface_hub[cli]\"\nhuggingface-cli download Wan-AI/Wan2.2-T2V-A14B --local-dir ./Wan2.2-T2V-A14B\n\n\nHere's an example of running Text-to-Video generation without prompt extension on a single GPU:\n\nsh\npython generate.py --task t2v-A14B --size 1280*720 --ckpt_dir ./Wan2.2-T2V-A14B --offload_model True --convert_model_dtype --prompt \"Two anthropomorphic cats in comfy boxing gear and bright gloves fight intensely on a spotlighted stage.\"\n\n\nFor multi-GPU inference or other generation tasks like Image-to-Video, Speech-to-Video, or Wan-Animate, please refer to the comprehensive documentation in the repository's README.\n\n## Why Use Wan2.2?\n\nWan2.2 stands out for its cutting-edge features and performance:\n\n*   **Mixture-of-Experts (MoE) Architecture:** Enhances model capacity and generation quality while maintaining efficient computational costs.\n*   **Cinematic-level Aesthetics:** Generates videos with precise and controllable cinematic styles, thanks to meticulously curated aesthetic data.\n*   **Complex Motion Generation:** Trained on significantly larger datasets, leading to superior generalization across motions, semantics, and aesthetics.\n*   **Efficient High-Definition TI2V:** The 5B model, powered by an advanced Wan2.2-VAE, supports 720P video generation at 24fps, even on consumer-grade GPUs like the RTX 4090.\n*   **Versatile Applications:** Supports Text-to-Video, Image-to-Video, Speech-to-Video, and Character Animation, making it suitable for a wide range of creative and industrial uses.\n*   **Open-Source:** Provides an accessible and powerful tool for researchers and developers in the generative AI community.\n\n## Links\n\n*   **GitHub Repository:** [Wan-Video/Wan2.2](https://github.com/Wan-Video/Wan2.2){:target=\"_blank\"}\n*   **Official Website:** [Wan.video](https://wan.video){:target=\"_blank\"}\n*   **Research Paper:** [arXiv:2503.20314](https://arxiv.org/abs/2503.20314){:target=\"_blank\"}\n*   **Hugging Face:** [Wan-AI](https://huggingface.co/Wan-AI/){:target=\"_blank\"}\n*   **Discord Community:** [Join Discord](https://discord.gg/AKNgpMK4Yj){:target=\"_blank\"}","metrics":{"detailViews":15,"githubClicks":17},"dates":{"published":null,"modified":"2025-10-29T00:01:02.000Z"}}