[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"project-72389":3},{"id":4,"name":5,"fullName":6,"owner":7,"repo":5,"description":8,"homepage":9,"htmlUrl":10,"language":11,"languages":10,"totalLinesOfCode":10,"stars":12,"forks":13,"watchers":14,"openIssues":15,"contributorsCount":16,"subscribersCount":16,"size":16,"stars1d":17,"stars7d":18,"stars30d":19,"stars90d":16,"forks30d":16,"starsTrendScore":20,"compositeScore":21,"rankGlobal":10,"rankLanguage":10,"license":22,"archived":23,"fork":23,"defaultBranch":24,"hasWiki":25,"hasPages":23,"topics":26,"createdAt":10,"pushedAt":10,"updatedAt":33,"readmeContent":34,"aiSummary":35,"trendingCount":16,"starSnapshotCount":16,"syncStatus":36,"lastSyncTime":37,"discoverSource":38},72389,"ComfyUI-nunchaku","nunchaku-ai\u002FComfyUI-nunchaku","nunchaku-ai","ComfyUI Plugin of Nunchaku","https:\u002F\u002Fnunchaku.tech\u002Fdocs\u002FComfyUI-nunchaku\u002F",null,"Python",2892,163,28,10,0,6,12,19,18,28.64,"Apache License 2.0",false,"main",true,[27,28,29,30,31,32],"comfyui","diffusion","flux","genai","mlsys","quantization","2026-06-12 02:03:02","\u003Cdiv align=\"center\" id=\"nunchaku_logo\">\n  \u003Cimg src=\"https:\u002F\u002Fhuggingface.co\u002Fdatasets\u002Fnunchaku-ai\u002Fcdn\u002Fresolve\u002Fmain\u002Flogo\u002Fv2\u002Fnunchaku-compact-transparent.png\" alt=\"logo\" width=\"220\">\u003C\u002Fimg>\n\u003C\u002Fdiv>\n\u003Ch3 align=\"center\">\n\u003Ca href=\"http:\u002F\u002Farxiv.org\u002Fabs\u002F2411.05007\">\u003Cb>Paper\u003C\u002Fb>\u003C\u002Fa> | \u003Ca href=\"https:\u002F\u002Fnunchaku.tech\u002Fdocs\u002FComfyUI-nunchaku\u002F\">\u003Cb>Docs\u003C\u002Fb>\u003C\u002Fa> | \u003Ca href=\"https:\u002F\u002Fhanlab.mit.edu\u002Fprojects\u002Fsvdquant\">\u003Cb>Website\u003C\u002Fb>\u003C\u002Fa> | \u003Ca href=\"https:\u002F\u002Fhanlab.mit.edu\u002Fblog\u002Fsvdquant\">\u003Cb>Blog\u003C\u002Fb>\u003C\u002Fa> | \u003Ca href=\"https:\u002F\u002Fdemo.nunchaku.tech\u002F\">\u003Cb>Demo\u003C\u002Fb>\u003C\u002Fa> | \u003Ca href=\"https:\u002F\u002Fhuggingface.co\u002Fnunchaku-ai\">\u003Cb>Hugging Face\u003C\u002Fb>\u003C\u002Fa> | \u003Ca href=\"https:\u002F\u002Fmodelscope.cn\u002Forganization\u002Fnunchaku-tech\">\u003Cb>ModelScope\u003C\u002Fb>\u003C\u002Fa>\n\u003C\u002Fh3>\n\n\u003Cdiv align=\"center\">\n  \u003Ca href=\"https:\u002F\u002Ftrendshift.io\u002Frepositories\u002F17711\" target=\"_blank\">\u003Cimg src=\"https:\u002F\u002Ftrendshift.io\u002Fapi\u002Fbadge\u002Frepositories\u002F17711\" alt=\"nunchaku-ai\u002Fnunchaku | Trendshift\" style=\"width: 120px; height: 26px;\" width=\"120\" height=\"26\"\u002F>\u003C\u002Fa>\n  \u003Ca href=https:\u002F\u002Fdiscord.gg\u002FWk6PnwX9Sm target=\"_blank\">\u003Cimg src=https:\u002F\u002Fimg.shields.io\u002Fbadge\u002Fdynamic\u002Fjson?url=https%3A%2F%2Fdiscord.com%2Fapi%2Finvites%2FWk6PnwX9Sm%3Fwith_counts%3Dtrue&query=%24.approximate_member_count&logo=discord&logoColor=white&label=Discord&color=green&suffix=%20total height=22px>\u003C\u002Fa>\n  \u003Ca href=https:\u002F\u002Fhuggingface.co\u002Fdatasets\u002Fnunchaku-ai\u002Fcdn\u002Fresolve\u002Fmain\u002Fnunchaku\u002Fassets\u002Fwechat.jpg target=\"_blank\">\u003Cimg src=https:\u002F\u002Fimg.shields.io\u002Fbadge\u002FWeChat-07C160?logo=wechat&logoColor=white height=22px>\u003C\u002Fa>\n  \u003Ca href=https:\u002F\u002Fdeepwiki.com\u002Fnunchaku-ai\u002FComfyUI-nunchaku target=\"_blank\">\u003Cimg src=https:\u002F\u002Fdeepwiki.com\u002Fbadge.svg height=22px>\u003C\u002Fa>\n\u003C\u002Fdiv>\n\nThis repository provides the ComfyUI plugin for [**Nunchaku**](https:\u002F\u002Fgithub.com\u002Fnunchaku-ai\u002Fnunchaku), an efficient inference engine for 4-bit neural networks quantized with [SVDQuant](http:\u002F\u002Farxiv.org\u002Fabs\u002F2411.05007). For the quantization library, check out [DeepCompressor](https:\u002F\u002Fgithub.com\u002Fnunchaku-ai\u002Fdeepcompressor).\n\nJoin our user groups on [**Discord**](https:\u002F\u002Fdiscord.gg\u002FWk6PnwX9Sm) and [**WeChat**](https:\u002F\u002Fhuggingface.co\u002Fdatasets\u002Fnunchaku-ai\u002Fcdn\u002Fresolve\u002Fmain\u002Fnunchaku\u002Fassets\u002Fwechat.jpg) for discussions—details [here](https:\u002F\u002Fgithub.com\u002Fnunchaku-ai\u002Fnunchaku\u002Fissues\u002F149). If you have any questions, run into issues, or are interested in contributing, feel free to share your thoughts with us!\n\n# Nunchaku ComfyUI Plugin\n\n![comfyui](https:\u002F\u002Fhuggingface.co\u002Fdatasets\u002Fnunchaku-ai\u002Fcdn\u002Fresolve\u002Fmain\u002FComfyUI-nunchaku\u002Fcomfyui.jpg)\n\n## News\n\n- **[2026-01-12]** 🚀 **v1.2.0 Released!** Enjoy a **20–30%** Z-Image performance boost, seamless **LoRA support** with native ComfyUI nodes, and INT4 support for **20-series GPUs**!\n- **[2025-12-26]** 🚀 **v1.1.0**: Support **4-bit [Tongyi-MAI\u002FZ-Image-Turbo](https:\u002F\u002Fhuggingface.co\u002FTongyi-MAI\u002FZ-Image-Turbo)**! Download on [Hugging Face](https:\u002F\u002Fhuggingface.co\u002Fnunchaku-ai\u002Fnunchaku-z-image-turbo) or [ModelScope](https:\u002F\u002Fmodelscope.cn\u002Fmodels\u002Fnunchaku-tech\u002Fnunchaku-z-image-turbo), and try it with this [workflow](.\u002Fexample_workflows\u002Fnunchaku-z-image-turbo.json)!\n- **[2025-09-24]** 🔥 Released **4-bit 4\u002F8-step Qwen-Image-Edit-2509 lightning** models at [Hugging Face](https:\u002F\u002Fhuggingface.co\u002Fnunchaku-ai\u002Fnunchaku-qwen-image-edit-2509)! Try them out with this [workflow](.\u002Fexample_workflows\u002Fnunchaku-qwen-image-edit-2509-lightning.json)!\n- **[2025-09-24]** 🔥 Released **4-bit Qwen-Image-Edit-2509**! Models are available on [Hugging Face](https:\u002F\u002Fhuggingface.co\u002Fnunchaku-ai\u002Fnunchaku-qwen-image-2509). Try them out with this [workflow](.\u002Fexample_workflows\u002Fnunchaku-qwen-image-edit-2509.json)!\n- **[2025-09-09]** 🔥 Released **4-bit Qwen-Image-Edit** together with the [4\u002F8-step Lightning](https:\u002F\u002Fhuggingface.co\u002Flightx2v\u002FQwen-Image-Lightning) variants! Models are available on [Hugging Face](https:\u002F\u002Fhuggingface.co\u002Fnunchaku-ai\u002Fnunchaku-qwen-image). Try them out with this [workflow](.\u002Fexample_workflows\u002Fnunchaku-qwen-image-edit.json)!\n\n\u003Cdetails>\n\u003Csummary>More\u003C\u002Fsummary>\n\n- **[2025-09-04]** 🚀 Official release of **Nunchaku v1.0.0**! Qwen-Image now supports **asynchronous offloading**, cutting Transformer VRAM usage to as little as **3 GiB** with no performance loss. You can also try our pre-quantized [4\u002F8-step Qwen-Image-Lightning](https:\u002F\u002Fhuggingface.co\u002Flightx2v\u002FQwen-Image-Lightning) models on [Hugging Face](https:\u002F\u002Fhuggingface.co\u002Fnunchaku-ai\u002Fnunchaku-qwen-image) or [ModelScope](https:\u002F\u002Fmodelscope.cn\u002Fmodels\u002Fnunchaku-tech\u002Fnunchaku-qwen-image).\n- **[2025-08-23]** 🚀 **v1.0.0** adds support for [Qwen-Image](https:\u002F\u002Fhuggingface.co\u002FQwen\u002FQwen-Image)! Check [this workflow](example_workflows\u002Fnunchaku-qwen-image.json) to get started. LoRA support is coming soon.\n- **[2025-07-17]** 📘 The official [**ComfyUI-nunchaku documentation**](https:\u002F\u002Fnunchaku.tech\u002Fdocs\u002FComfyUI-nunchaku\u002F) is now live! Explore comprehensive guides and resources to help you get started.\n- **[2025-06-29]** 🔥 **v0.3.3** now supports [FLUX.1-Kontext-dev](https:\u002F\u002Fhuggingface.co\u002Fblack-forest-labs\u002FFLUX.1-Kontext-dev)! Download the quantized model from [Hugging Face](https:\u002F\u002Fhuggingface.co\u002Fnunchaku-ai\u002Fnunchaku-flux.1-kontext-dev) or [ModelScope](https:\u002F\u002Fmodelscope.cn\u002Fmodels\u002Fnunchaku-tech\u002Fnunchaku-flux.1-kontext-dev) and use this [workflow](.\u002Fexample_workflows\u002Fnunchaku-flux.1-kontext-dev.json) to get started.\n- **[2025-06-11]** Starting from **v0.3.2**, you can now **easily install or update the [Nunchaku](https:\u002F\u002Fgithub.com\u002Fnunchaku-ai\u002Fnunchaku) wheel** using this [workflow](https:\u002F\u002Fgithub.com\u002Fnunchaku-ai\u002FComfyUI-nunchaku\u002Fblob\u002Fmain\u002Fexample_workflows\u002Finstall_wheel.json)!\n- **[2025-06-07]** 🚀 **Release Patch v0.3.1!** We bring back **FB Cache** support and fix **4-bit text encoder loading**. PuLID nodes are now optional and won’t interfere with other nodes. We've also added a **NunchakuWheelInstaller** node to help you install the correct [Nunchaku](https:\u002F\u002Fgithub.com\u002Fnunchaku-ai\u002Fnunchaku) wheel.\n- **[2025-06-01]** 🚀 **Release v0.3.0!** This update adds support for multiple-batch inference, [**ControlNet-Union-Pro 2.0**](https:\u002F\u002Fhuggingface.co\u002FShakker-Labs\u002FFLUX.1-dev-ControlNet-Union-Pro-2.0) and initial integration of [**PuLID**](https:\u002F\u002Fgithub.com\u002FToTheBeginning\u002FPuLID). You can now load Nunchaku FLUX models as a single file, and our upgraded [**4-bit T5 encoder**](https:\u002F\u002Fhuggingface.co\u002Fnunchaku-ai\u002Fnunchaku-t5) now matches **FP8 T5** in quality!\n- **[2025-04-16]** 🎥 Released tutorial videos in both [**English**](https:\u002F\u002Fyoutu.be\u002FYHAVe-oM7U8?si=cM9zaby_aEHiFXk0) and [**Chinese**](https:\u002F\u002Fwww.bilibili.com\u002Fvideo\u002FBV1BTocYjEk5\u002F?share_source=copy_web&vd_source=8926212fef622f25cc95380515ac74ee) to assist installation and usage.\n- **[2025-04-09]** 📢 Published the [April roadmap](https:\u002F\u002Fgithub.com\u002Fnunchaku-ai\u002Fnunchaku\u002Fissues\u002F266) and an [FAQ](https:\u002F\u002Fgithub.com\u002Fnunchaku-ai\u002Fnunchaku\u002Fdiscussions\u002F262) to help the community get started and stay up to date with Nunchaku’s development.\n- **[2025-04-05]** 🚀 **Release v0.2.0!** This release introduces [**multi-LoRA**](example_workflows\u002Fnunchaku-flux.1-dev.json) and [**ControlNet**](example_workflows\u002Fnunchaku-flux.1-dev-controlnet-union-pro.json) support, with enhanced performance using FP16 attention and First-Block Cache. We've also added [**20-series GPU**](examples\u002Fflux.1-dev-turing.py) compatibility and official workflows for [FLUX.1-redux](example_workflows\u002Fnunchaku-flux.1-redux-dev.json)!\n\n\u003C\u002Fdetails>\n\n## Getting Started\n\n- [Installation Guide](https:\u002F\u002Fnunchaku.tech\u002Fdocs\u002FComfyUI-nunchaku\u002Fget_started\u002Finstallation.html)\n- [Usage Tutorial](https:\u002F\u002Fnunchaku.tech\u002Fdocs\u002FComfyUI-nunchaku\u002Fget_started\u002Fusage.html)\n- [Example Workflows](https:\u002F\u002Fnunchaku.tech\u002Fdocs\u002FComfyUI-nunchaku\u002Fworkflows\u002Ftoc.html)\n- [Node Reference](https:\u002F\u002Fnunchaku.tech\u002Fdocs\u002FComfyUI-nunchaku\u002Fnodes\u002Ftoc.html)\n- [API Reference](https:\u002F\u002Fnunchaku.tech\u002Fdocs\u002FComfyUI-nunchaku\u002Fapi\u002Ftoc.html)\n- [Custom Model Quantization: DeepCompressor](https:\u002F\u002Fgithub.com\u002Fmit-han-lab\u002Fdeepcompressor)\n- [Contribution Guide](https:\u002F\u002Fnunchaku.tech\u002Fdocs\u002FComfyUI-nunchaku\u002Fdeveloper\u002Fcontribution_guide.html)\n- [Frequently Asked Questions](https:\u002F\u002Fnunchaku.tech\u002Fdocs\u002Fnunchaku\u002Ffaq\u002Ffaq.html)\n\n## Star History\n\n[![Star History Chart](https:\u002F\u002Fapi.star-history.com\u002Fsvg?repos=nunchaku-ai\u002FComfyUI-nunchaku&type=Date)](https:\u002F\u002Fwww.star-history.com\u002F#nunchaku-ai\u002FComfyUI-nunchaku&Date)\n","ComfyUI-nunchaku 是一个为 Nunchaku 设计的 ComfyUI 插件，Nunchaku 是一种高效的 4 位神经网络推理引擎，使用 SVDQuant 进行量化。该插件支持 Z-Image 模型性能提升、LoRA 支持以及对 20 系列 GPU 的 INT4 支持等核心功能。项目采用 Python 编写，并且遵循 Apache License 2.0 开源许可协议。它特别适用于需要高效执行深度学习模型推理的应用场景，尤其是在资源受限的环境中，比如移动设备或边缘计算平台。此外，通过加入 Discord 和微信用户群可以获得更多的技术支持和交流机会。",2,"2026-06-11 03:41:50","high_star"]