---
title: awesome-comfyui-models
canonical_url: "https://www.modelscope.cn/models/siliconflow/awesome-comfyui-models"
md_url: "https://www.modelscope.cn/models/siliconflow/awesome-comfyui-models.md"
repository: siliconflow/awesome-comfyui-models
last_updated: 2026-10-29
pipeline_tag: any-to-any
tasks:
  - any-to-any
model_type:
  - neo_chat
architectures:
  - NEOChatModel
parameters: 1175.4B
tensor_type:
  - F32
  - I64
  - F16
  - F8_E4M3
  - I8
  - BF16
  - U8
  - I32
  - BOOL
library_name:
  - safetensors
  - onnx
  - diffusers
  - gguf
  - pytorch
frameworks:
  - pytorch
downloads: 12286884
stars: 15
tags:
  - gguf
---

# awesome-comfyui-models

> awesome-comfyui-models - siliconflow 在 ModelScope 开源的模型。注意：使用模型时，请参考 URL 标注的模型来源，并遵守 License 和 Commercial Use 中标注的协议（以原模型协议为准）。如有异议，请联系我们做删除。

- **Repository**: siliconflow/awesome-comfyui-models
- **Tasks**: any-to-any
- **Parameters**: 1175.4B
- **Tags**: gguf
- **Downloads**: 12286884
- **Stars**: 15
- **Last updated**: 2026-10-29

Source: https://www.modelscope.cn/models/siliconflow/awesome-comfyui-models

---

### 模型介绍

注意：使用模型时，请参考 URL 标注的模型来源，并遵守 License 和 Commercial Use 中标注的协议（以原模型协议为准）。如有异议，请联系我们做删除。

| No. | WeightID | Notes | ComfyUIPath | URL | Hash | License | Commercial Use |
|-----|----------|-------|-------------|-----|------|---------|----------|
| 1 | flux1-dev-fp8.safetensors | comfyui官方提供的的flux_fp8 模型, 融合了 Clip 及 VAE | checkpoints | [Link](https://huggingface.co/Comfy-Org/flux1-dev/blob/ca7d619b1bcd7156ca897f55efb2370eba3d9e20/flux1-dev-fp8.safetensors) | SHA256: 8e91b68084b53a7fc44ed2a3756d821e355ac1a7b6fe29be760c1db532f3d88a | FLUX.1 [dev] Non-Commercial License | no |
| 2 | sd3.5_large.safetensors | SD3.5 large 官方模型 | checkpoints | [Link](https://huggingface.co/stabilityai/stable-diffusion-3.5-large/blob/4c4df3fa464561ba37e7c1a2c25da50f181f8206/sd3.5_large.safetensors) | SHA256: ffef7a279d9134626e6ce0d494fba84fc1c7e720b3c7df2d19a09dc3796d8f93 | stabilityai-ai-community | no |
| 3 | SUPIR-v0Q.ckpt | 用于图像超分辨率的高性能模型，适用于8K图像放大任务 | checkpoints | [Link](https://huggingface.co/camenduru/SUPIR/blob/97b24ec4d42bbdf6b3c5a8d701f78ac67aacba04/SUPIR-v0Q.ckpt) | SHA256: d7f418398bb024d0d3c779c4ee4e9f171eb072306093f5bcfb2bf096aa2738f8 | No license is provided | no |
| 4 | Juggernaut-XI-byRunDiffusion.safetensors | 基于SDXL优化的人像增强模型，适用于高质量写实风格图像生成 | checkpoints | [Link](https://huggingface.co/RunDiffusion/Juggernaut-XI-v11/blob/c5ba39535f38be10a7b617ef6e294088769b2dbc/Juggernaut-XI-byRunDiffusion.safetensors) | SHA256: 33e58e86686f6b386c526682b5da9228ead4f91d994abd4b053442dc5b42719e | CC-BY-NC-ND-4.0 License | no |
| 5 | dreamshaper_8.safetensors | DreamShaper 美术和插画模型，来自于 https://civitai.com/models/4384?modelVersionId=128713 | checkpoints | [Link](https://huggingface.co/digiplay/DreamShaper_8/blob/eaf8bf3c1ea3ed83ddfc953dce79fd75307676a9/dreamshaper_8.safetensors) | SHA256: 879db523c30d3b9017143d56705015e15a2cb5628762c11d086fed9538abd7fd | CreativeML Open RAIL -M License | yes |
| 6 | dreamshaperXL_lightningDPMSDE.safetensors | inpaint anything 官方模型 | checkpoints | [Link](https://huggingface.co/gingerlollipopdx/ModelsXL/blob/de865e6bf0a37463c9750ea67a180e2766bb7e55/dreamshaperXL_lightningDPMSDE.safetensors) | SHA256: fdbe56354b8f876b736f24d3ad867ecd4140c019f57642ec8ddd878088d44f64 | No license is provided | no |
| 7 | albedobaseXL_v13.safetensors | Better Face Swap工作流依赖的SDXL 骨干模型 | checkpoints | [Link](https://huggingface.co/frankjoshua/albedobaseXL_v13/blob/bcd5fa3c8d2efed783376792c61c78115113be90/albedobaseXL_v13.safetensors) | SHA256: a40d817f46acc25c88ee0b6f6efa474004cea3fa1868e09a13192342c5bf4df9 | No license is provided | no |
| 8 | sd_xl_base_1.0.safetensors | 文本生成图像的主模型, 使用OpenCLIP-ViT/G和CLIP-ViT/L两个文本编码器 | checkpoints | [Link](https://huggingface.co/stabilityai/stable-diffusion-xl-base-1.0/blob/f298da3c058bd8f1f1c62f3ecfa775244a243897/sd_xl_base_1.0.safetensors) | SHA256: 31e35c80fc4829d14f90153f4c74cd59c90b779f6afe05a74cd6120b893f7e5b | CreativeML Open RAIL++-M License | yes |
| 9 | sd-v1-5-inpainting.ckpt | SD 1.5 的图像修复专用模型，支持基于掩码的局部重绘 | checkpoints | [Link](https://huggingface.co/stable-diffusion-v1-5/stable-diffusion-inpainting/blob/8a4288a76071f7280aedbdb3253bdb9e9d5d84bb/sd-v1-5-inpainting.ckpt) | SHA256: c6bbc15e3224e6973459ba78de4998b80b50112b0ae5b5c67113d56b4e366b19 | Responsible AI License | yes |
| 10 | v1-5-pruned-emaonly.ckpt | SD 1.5 剪枝版本 | checkpoints | [Link](https://huggingface.co/stable-diffusion-v1-5/stable-diffusion-v1-5/blob/f03de327dd89b501a01da37fc5240cf4fdba85a1/v1-5-pruned-emaonly.ckpt) | SHA256: 6ce0161689b3853acaa03779ec93eafe75a02f4ced659bee03f50797806fa2fa | Responsible AI License | yes |
| 11 | svd_xt.safetensors | 基于图像条件反射生成短视频的潜在扩散模型 | checkpoints | [Link](https://huggingface.co/stabilityai/stable-video-diffusion-img2vid-xt/blob/9e43909513c6714f1bc78bcb44d96e733cd242aa/svd_xt.safetensors) | SHA256: b2652c23d64a1da5f14d55011b9b6dce55f2e72e395719f1cd1f8a079b00a451 | CreativeML Open RAIL++-M License | yes |
| 12 | realisticVisionV20_v20.safetensors | 基于 Stable Diffusion 1.5 架构的写实风格图像生成模型，根据文字提示生成高写实图像 | checkpoints | [Link](https://huggingface.co/ckpt/realistic-vision-v20/blob/bde2dbc9931776a0552fbb27ed1b562e352feab2/realisticVisionV20_v20.safetensors) | SHA256: e6415c48928aedc891237556125e56ebd0bb3d9f0982d469ebb719d1117dfb92 | FLUX.1 [dev] Non-Commercial License | no |
| 13 | dreamshaperXL_sfwLightningDPMSDE.safetensors | SDXL 模型 DreamShaper XL Lightning 版本为 3–6 步采样定制，可较快生成高质量的动漫风格图像 | checkpoints | [Link](https://civitai.com/models/112902?modelVersionId=354657) | SHA256: FDBE56354B8F876B736F24D3AD867ECD4140C019F57642EC8DDD878088D44F64 | CreativeML Open RAIL++-M License | yes |
| 14 | epicRealism.safetensors | 基于 SD1.5 的文生图模型，专注于生成写实风格的人像与场景图像 | checkpoints | [Link](https://civitai.com/models/25694/epicrealism) | SHA256: f42428d476f50e72158c0b62aadf63a6c664251bbb7ec4f9cf6dc1576ffe50bc | CreativeML Open RAIL -M License | yes |
| 15 | hunyuan3d-dit-v2.safetensors | Tencent发布的开源3D资产生成模型，能够通过文本或图像生成具有高分辨率纹理图的高保真3D模型，仅允许单视图输入 | checkpoints | [Link](https://huggingface.co/tencent/Hunyuan3D-2/blob/ca77459c78ddce265434f2d351e77b4574ba78eb/hunyuan3d-dit-v2-0/model.fp16.safetensors?download=true) | SHA256: 360bc281fc956d4acac0c3d36d5ec0ebf8cdddbf4b8892e894d12419388d479b | tencent-hunyuan-community | N/A |
| 16 | hunyuan3d-dit-v2-mv.safetensors | Tencent发布的开源3D资产生成模型，能够通过文本或图像生成具有高分辨率纹理图的高保真3D模型，允许多视图输入 | checkpoints | [Link](https://huggingface.co/tencent/Hunyuan3D-2mv/blob/08766051fa711c6ef5caf86b97e50304fdfcf0ef/hunyuan3d-dit-v2-mv/model.fp16.safetensors?download=true) | SHA256: d36f5881bcdc56726b73e517cd444c13c60732431622da7268145355c8d38e9c | tencent-hunyuan-community | N/A |
| 17 | ace_step_v1_3.5b.safetensors | ACE-Step 是由中国团队 StepFun 和 ACE Studio 联合开发的开源基础音乐生成模型，旨在为音乐创作者提供高效、灵活和高质量的音乐生成和编辑工具 | checkpoints | [Link](https://huggingface.co/Comfy-Org/ACE-Step_ComfyUI_repackaged/blob/e6d4f4e0ba17b24b35a3fc3b83c8e1c83b7faf9c/all_in_one/ace_step_v1_3.5b.safetensors) | SHA256: f07cad74c4adce52ca14ca1bdf74cf3c14cbafb0823b95eca4459467fa369f40 | Apache-2.0 License | yes |
| 18 | stable_audio_open_1.0.safetensors | Stable Audio Open是一个基于文本的音频生成模型，能够根据简单的文本提示生成长达47秒的高质量音频数据 | checkpoints | [Link](https://huggingface.co/stabilityai/stable-audio-open-1.0/blob/7b1754f6e720548869ad4a25891734eb63e3a3cc/model.safetensors) | 7b20458a071231aaf32613b6fbc7945f28f34dbba4f295bb49bad56f5f66b57e | stable-audio-community | N/A |
| 19 | chord_v1.safetensors | 基于生成纹理图像的 PBR 材质估计模型 | checkpoints | [Link](https://huggingface.co/Ubisoft/ubisoft-laforge-chord/blob/27b902c851478bb709108bb49dd95e093fb0821e/chord_v1.safetensors) | SHA256: b56f48857f1d84f0d051284d0c1f3ac58ecc6e86fba15c4849789f3807dd0cc2 | ubisoft-machine-learning-license-research-only-copyleft | N/A |
| 20 | animagine-xl-3.1.safetensors | 动漫主题生图模型 | checkpoints/XL | [Link](https://huggingface.co/cagliostrolab/animagine-xl-3.1/blob/b92a35d36f149c6187e916c3cf74961066121092/animagine-xl-3.1.safetensors) | SHA256: e3c47aedb06418c6c331443cd89f2b3b3b34b7ed2102a3d4c4408a8d35aad6b0 | Fair AI Public License 1.0-SD | yes |
| 21 | vae-ft-mse-840000-ema-pruned.safetensors | 基于 Stable Diffusion 1.x 系列优化的 变分自编码器(VAE)模型, 实际应用中常用的VAE, 接近于实际拍摄, 提升图像饱和度和细节质量 | vae | [Link](https://huggingface.co/stabilityai/sd-vae-ft-mse-original/blob/629b3ad3030ce36e15e70c5db7d91df0d60c627f/vae-ft-mse-840000-ema-pruned.safetensors) | SHA256: 735e4c3a447a3255760d7f86845f09f937809baa529c17370d83e4c3758f3c75 | MIT License | yes |
| 22 | sdxl_vae.safetensors | Stable Diffusion XL 的官方 VAE 解码器 | vae | [Link](https://huggingface.co/stabilityai/sdxl-vae/blob/c1b803cdacd792f788384449b8e1d539da4297ad/sdxl_vae.safetensors) | SHA256: 63aeecb90ff7bc1c115395962d3e803571385b61938377bc7089b36e81e92e2e | MIT License | yes |
| 23 | ae.safetensors | FLUX.1 系列模型的自动编码器 | vae | [Link](https://huggingface.co/black-forest-labs/FLUX.1-dev/blob/e8be9d277dbb13b0f8d31eb36e2efcfe4b3a880c/ae.safetensors) | SHA256: afc8e28272cd15db3919bacdb6918ce9c1ed22e96cb12c4d5ed0fba823529e38 | FLUX.1 [dev] Non-Commercial License | no |
| 24 | wan2.2_vae.safetensors | Wan2.2 模型专用的 VAE 权重文件 | vae | [Link](https://huggingface.co/Comfy-Org/Wan_2.2_ComfyUI_Repackaged/blob/8441d066add15eae8d84f42aa6d9c45417973ce6/split_files/vae/wan2.2_vae.safetensors) | SHA256: e40321bd36b9709991dae2530eb4ac303dd168276980d3e9bc4b6e2b75fed156 | Apache-2.0 License | yes |
| 25 | wan_2.1_vae.safetensors | Wan2.1 模型专用的 VAE 权重文件 | vae | [Link](https://huggingface.co/Comfy-Org/Wan_2.2_ComfyUI_Repackaged/blob/0b080e63b27fd0ba11fbee3597e41a23493358eb/split_files/vae/wan_2.1_vae.safetensors) | SHA256: 2fc39d31359a4b0a64f55876d8ff7fa8d780956ae2cb13463b0223e15148976b | No license is provided | no |
| 26 | qwen_image_vae.safetensors | Qwen 图像解码的 VAE 模型，在图像生成和后期重构中维持高准确性 | vae | [Link](https://huggingface.co/Comfy-Org/Qwen-Image_ComfyUI/blob/dfe60a0d63f0b946628080f070978594983b8b6e/split_files/vae/qwen_image_vae.safetensors) | SHA256: a70580f0213e67967ee9c95f05bb400e8fb08307e017a924bf3441223e023d1f | Apache-2.0 License | yes |
| 27 | hunyuan_image_2.1_vae_fp16.safetensors | HunyuanImage 2.1 VAE 模型权重文件 | vae | [Link](https://huggingface.co/Comfy-Org/HunyuanImage_2.1_ComfyUI/blob/45466c4d65dbe5996aff90b3420d452a522db9ef/split_files/vae/hunyuan_image_2.1_vae_fp16.safetensors) | SHA256: f2ae19863609206196b5e3a86bfd94f67bd3866f5042004e3994f07e3c93b2f9 | No license is provided | no |
| 28 | hunyuanvideo15_vae_fp16.safetensors | HunyuanVideo 1.5 专用 VAE，FP16 精度，用于视频潜空间的编码与解码 | vae | [Link](https://huggingface.co/Comfy-Org/HunyuanVideo_1.5_repackaged/resolve/main/split_files/vae/hunyuanvideo15_vae_fp16.safetensors) | SHA256: e7c3091949c27e2d55ae6d5df917b99dadfebbf308e5a50d0ade0d16c90297ae | tencent-hunyuan-community | N/A |
| 29 | flux2-vae.safetensors | FLUX.2 专用 VAE 自动编码器，用于编码/解码潜空间，提升生成图像的细节与还原度 | vae | [Link](https://huggingface.co/Comfy-Org/flux2-dev/resolve/main/split_files/vae/flux2-vae.safetensors) | No hash is provided | Apache-2.0 License | yes |
| 30 | qwen_image_layered_vae.safetensors | Qwen-Image-Layered 专用 VAE 自动编码器，来自 Comfy-Org/Qwen-Image-Layered_ComfyUI  | vae | [Link](https://huggingface.co/Comfy-Org/Qwen-Image-Layered_ComfyUI/resolve/main/split_files/vae/qwen_image_layered_vae.safetensors) | SHA256: d4fed3a99808eb21d2ee40c8243d92c2999e66bcbc6333b839a5b9c653db4a3e | Apache-2.0 License | yes |
| 31 | flux1-dev.safetensors | flux dev 骨干模型，fp16 格式 | unet | [Link](https://huggingface.co/black-forest-labs/FLUX.1-dev/blob/0ef5fff789c832c5c7f4e127f94c8b54bbcced44/flux1-dev.safetensors) | SHA256: 4610115bb0c89560703c892c59ac2742fa821e60ef5871b33493ba544683abd7 | FLUX.1 [dev] Non-Commercial License | no |
| 32 | flux1-dev-fp8-e4m3fn.safetensors | flux dev 骨干模型的fp8量化版本,量化类型为e4m3 | unet | [Link](https://huggingface.co/Kijai/flux-fp8/blob/90b5f6edd6042e9b8e51e33164382293d627cd8f/flux1-dev-fp8-e4m3fn.safetensors) | SHA256: 47d8dbdc6da6e996ff31567476c0a549afd1a99554b20af2be57f0ccfa93aa36 | FLUX.1 [dev] Non-Commercial License | no |
| 33 | flux1-schnell-fp8-e4m3fn.safetensors | flux1-schnell fp8量化模型 | unet | [Link](https://huggingface.co/Kijai/flux-fp8/blob/6636d43abd7be5d9d6d4342cb77b3fffb1857551/flux1-schnell-fp8-e4m3fn.safetensors) | SHA256: bbdfba27fed8ff3be237523fb37b83821a6c4bbaa1db43ef9288767d0e4042fb | Apache-2.0 License | yes |
| 34 | iclight_sd15_fc_unet_ldm.safetensors | ic‑light 项目中用于稳定扩散 1.5 的默认文本条件光照重塑模型，可根据文本与前景图生成一致光照效果 | unet | [Link](https://huggingface.co/lllyasviel/ic-light/blob/f5950d474dc0cd5bb5a0c66189534cd13f28eb70/iclight_sd15_fc.safetensors) | SHA256: a033fbaaa2f3f7859fa6a4477ee63ebbf9c116bf3569d5811856d2807f3468cd | Apache-2.0 License | yes |
| 35 | flux1-dev-kontext_fp8_scaled.safetensors | 低显存高效推理用的 FP8 量化版 Flux1 Kontext UNet 权重文件（带缩放因子 scaled） | unet | [Link](https://huggingface.co/Comfy-Org/flux1-kontext-dev_ComfyUI/blob/c215e51f94373df8914054c98ba67575c0c2b157/split_files/diffusion_models/flux1-dev-kontext_fp8_scaled.safetensors) | SHA256: 630ba795ec64283b4230ea23cf79406c2c68b7c578229ed139f30043eadb30a2 | FLUX.1 [dev] Non-Commercial License | no |
| 36 | omnigen2_fp16.safetensors | 多模态图像生成模型 OmniGen2 | unet | [Link](https://huggingface.co/Comfy-Org/Omnigen2_ComfyUI_repackaged/blob/4876f2222e35e269029e8d72aaff5b2aaaf73e1b/split_files/diffusion_models/omnigen2_fp16.safetensors) | SHA256: 60dbde45107762d164bac463e1cf365e074b377fa843dc90cb2985fb211cd4de | Apache-2.0 License | yes |
| 37 | flux1-kontext-dev.safetensors | Black Forest Labs提供的Flux1 Kontext fp16格式权重文件 | unet | [Link](https://huggingface.co/black-forest-labs/FLUX.1-Kontext-dev/blob/0018ada/flux1-kontext-dev.safetensors) | SHA256: 843a26dc765d3105dba081c30bce7b14c65b0988f9e8d14e9fbc8856a6deebd5 | FLUX.1 [dev] Non-Commercial License | no |
| 38 | neta-lumina-v1.0.safetensors | Neta-art 团队开发的高质量动漫风格的文本生成图像模型 | unet | [Link](https://huggingface.co/neta-art/Neta-Lumina/blob/e8d149e8b2c77ff35dd26b0cc7c3e677c985977d/Unet/neta-lumina-v1.0.safetensors) | SHA256: 6e57bb17e7618aa8bb169c06e131f3454c4c867e33b8721f2d06d2c146de7834 | Apache-2.0 License | yes |
| 39 | wan2.2_t2v_low_noise_14B_fp8_scaled.safetensors | Wan2.2 模型中用于文本生成视频任务的低噪声优化版本，140 亿参数，从文本描述生成高质量的视频，具有电影级的美学控制和精确的语义遵从性，采用缩放后的 FP8 精度 | unet | [Link](https://huggingface.co/Comfy-Org/Wan_2.2_ComfyUI_Repackaged/blob/0b080e63b27fd0ba11fbee3597e41a23493358eb/split_files/diffusion_models/wan2.2_t2v_low_noise_14B_fp8_scaled.safetensors) | SHA256: e71b96d7c82e638694c5e7fb98fac4bfb0e4ddc5fbbb4b1df40da8f0f1278a97 | Apache-2.0 License | yes |
| 40 | wan2.2_t2v_low_noise_14B_fp16.safetensors | Wan2.2 模型中用于文本生成视频任务的低噪声优化版本，140 亿参数，从文本描述生成高质量的视频，具有电影级的美学控制和精确的语义遵从性, FP16 精度 | unet | [Link](https://huggingface.co/Comfy-Org/Wan_2.2_ComfyUI_Repackaged/blob/7c19a08c82704c754159cdff9e934b23646c7a0c/split_files/diffusion_models/wan2.2_t2v_low_noise_14B_fp16.safetensors) | SHA256: 431d1613ffa809ae1f735b661a01788c6d74991f51efd01f45d5aee955ccd224 | Apache-2.0 License | yes |
| 41 | wan2.2_t2v_high_noise_14B_fp8_scaled.safetensors | Wan2.2 模型中用于文本生成视频任务的高噪声优化版本，采用缩放后的 FP8 精度 | unet | [Link](https://huggingface.co/Comfy-Org/Wan_2.2_ComfyUI_Repackaged/blob/0b080e63b27fd0ba11fbee3597e41a23493358eb/split_files/diffusion_models/wan2.2_t2v_high_noise_14B_fp8_scaled.safetensors) | SHA256: cad711ae211c8b23455ec68cd6a190a33a3d874234a77eb57266d73f8f0e6c9f | Apache-2.0 License | yes |
| 42 | wan2.2_t2v_high_noise_14B_fp16.safetensors | Wan2.2 模型中用于文本生成视频任务的高噪声优化版本，FP16 精度 | unet | [Link](https://huggingface.co/Comfy-Org/Wan_2.2_ComfyUI_Repackaged/blob/7c19a08c82704c754159cdff9e934b23646c7a0c/split_files/diffusion_models/wan2.2_t2v_high_noise_14B_fp16.safetensors) | SHA256: c793e1515320fc8883ac8f97e2a45270c35edd8c4e334a6d42d3e95455bd8da2 | Apache-2.0 License | yes |
| 43 | wan2.2_i2v_low_noise_14B_fp8_scaled.safetensors | Wan2.2 模型用于图像生成视频任务的低噪声阶段权重文件，140 亿参数，将静态图像转换为动态视频，保持内容的一致性和流畅的动态过程，采用缩放后的 FP8 精度 | unet | [Link](https://huggingface.co/Comfy-Org/Wan_2.2_ComfyUI_Repackaged/blob/0b080e63b27fd0ba11fbee3597e41a23493358eb/split_files/diffusion_models/wan2.2_i2v_low_noise_14B_fp8_scaled.safetensors) | SHA256: 5471a457b6ac404202a5fbe6c11595a3d5641fc766b00f38763f72303fffc21e | Apache-2.0 License | yes |
| 44 | wan2.2_i2v_low_noise_14B_fp16.safetensors | Wan2.2 模型用于图像生成视频任务的低噪声阶段权重文件，140 亿参数，将静态图像转换为动态视频，保持内容的一致性和流畅的动态过程，FP16 精度 | unet | [Link](https://huggingface.co/Comfy-Org/Wan_2.2_ComfyUI_Repackaged/blob/7c19a08c82704c754159cdff9e934b23646c7a0c/split_files/diffusion_models/wan2.2_i2v_low_noise_14B_fp16.safetensors) | SHA256: edb89340c8a6fbf1a70e76a839ae01eaf7d289f05ea1ebd6b1a3fc6f533826e9 | Apache-2.0 License | yes |
| 45 | wan2.2_i2v_high_noise_14B_fp8_scaled.safetensors | Wan2.2 模型用于图像生成视频任务的高噪声阶段权重文件，采用缩放后的 FP8 精度 | unet | [Link](https://huggingface.co/Comfy-Org/Wan_2.2_ComfyUI_Repackaged/blob/0b080e63b27fd0ba11fbee3597e41a23493358eb/split_files/diffusion_models/wan2.2_i2v_high_noise_14B_fp8_scaled.safetensors) | SHA256: 6122e79d55e0f235698d11d657f3b196c5273c830da00b2b013c5a048d5e6a42 | Apache-2.0 License | yes |
| 46 | wan2.2_i2v_high_noise_14B_fp16.safetensors | Wan2.2 模型用于图像生成视频任务的高噪声阶段权重文件，FP16 精度 | unet | [Link](https://huggingface.co/Comfy-Org/Wan_2.2_ComfyUI_Repackaged/blob/7c19a08c82704c754159cdff9e934b23646c7a0c/split_files/diffusion_models/wan2.2_i2v_high_noise_14B_fp16.safetensors) | SHA256: c21c21efa368d529a982a7d89dd11d86608b5937cae9a125d0ffe5491918a100 | Apache-2.0 License | yes |
| 47 | qwen_image_bf16.safetensors | 200亿参数 MMDiT 图像基础模型，支持中英复杂文本高保真渲染与一致性编辑 (bf16 版本） | unet | [Link](https://huggingface.co/Comfy-Org/Qwen-Image_ComfyUI/blob/dfe60a0d63f0b946628080f070978594983b8b6e/split_files/diffusion_models/qwen_image_bf16.safetensors) | SHA256: d08fb5d68026c0d87325f9a7b3ad6454061113a2bc73cc883114dae172937ae7 | Apache-2.0 License | yes |
| 48 | qwen_image_fp8_e4m3fn.safetensors | 200亿参数多模态扩散 Transformer 图像基础模型，支持中英复杂文本高保真渲染与一致性编辑（FP8 量化版本） | unet | [Link](https://huggingface.co/Comfy-Org/Qwen-Image_ComfyUI/blob/a4875863818e93fdb65f366d1d7a201581d05eeb/split_files/diffusion_models/qwen_image_fp8_e4m3fn.safetensors) | SHA256: 98763a127701eb6fb59096f7742cb3aa7d64ed510b9f4e882d8351f8176e3ce3 | Apache-2.0 License | yes |
| 49 | flux1-krea-dev.safetensors | Black Forest Labs 与 Krea 合作发布的文生图模型，生成更真实的图像，不包含过饱和纹理 | unet | [Link](https://huggingface.co/black-forest-labs/FLUX.1-Krea-dev/blob/f6315581b7cddd450b9aba72b4e9ccf8b6580dc1a6b9538aff43ee26a1a3b6c2/flux1-krea-dev.safetensors) | SHA256: f6315581b7cddd450b9aba72b4e9ccf8b6580dc1a6b9538aff43ee26a1a3b6c2 | FLUX.1 [dev] Non-Commercial License | no |
| 50 | flux1-depth-dev.safetensors | 120 亿参数 Rectified Flow Transformer 模型，根据文本描述生成图像，保持输入图像的结构特征，Depth版本通过深度图提取技术保持源图像的空间结构 | unet | [Link](https://huggingface.co/black-forest-labs/FLUX.1-Depth-dev/blob/e4b768dc0a9ca9837f4f026cf2c13e0de5f5c095/flux1-depth-dev.safetensors) | SHA256: 41360d1662f44ca45bc1b665fe6387e91802f53911001630d970a4f8be8dac21 | FLUX.1 [dev] Non-Commercial License | no |
| 51 | flux1-canny-dev.safetensors | 120 亿参数 Rectified Flow Transformer 模型，根据文本描述生成图像，保持输入图像的结构特征，Canny版本使用边缘检测技术保留源图像的结构特征 | unet | [Link](https://huggingface.co/black-forest-labs/FLUX.1-Canny-dev/blob/76514f5c24032730f3b23a4e42a02d2809b34d80/flux1-canny-dev.safetensors) | SHA256: 996876670169591cb412b937fbd46ea14cbed6933aef17c48a2dcd9685c98cdb | FLUX.1 [dev] Non-Commercial License | no |
| 52 | flux1-fill-dev.safetensors | Black Forest Labs 发布的约 120 亿参数的开源 Dev 版本模型，专用于基于文本提示对图片中的区域进行高质量修复（inpainting）和扩展（outpainting） | unet | [Link](https://huggingface.co/black-forest-labs/FLUX.1-Fill-dev/blob/9366c06b7b9111aa2e872a235eda1a0c37134695/flux1-fill-dev.safetensors) | SHA256: 03e289f530df51d014f48e675a9ffa2141bc003259bf5f25d75b957e920a41ca | FLUX.1 [dev] Non-Commercial License | no |
| 53 | qwen_image_edit_fp8_e4m3fn.safetensors | Qwen 图像编辑的 FP8 模型，专注于图像编辑任务，支持精确文本编辑 | unet | [Link](https://huggingface.co/Comfy-Org/Qwen-Image-Edit_ComfyUI/blob/3fdaf0f61aaeff8e0b22463bd7f506c68175fa4c/split_files/diffusion_models/qwen_image_edit_fp8_e4m3fn.safetensors) | SHA256: 393c6743d1de2e9031b5197027b36116f2096958ccc0223526d34e1860266021 | Apache-2.0 License | yes |
| 54 | hunyuanimage2.1.safetensors | Tencent-Hunyuan 发布的 HunyuanImage-2.1 文本到图像扩散模型的权重文件，用于模型在 2K 分辨率生成高质量图像 | unet | [Link](https://huggingface.co/Comfy-Org/HunyuanImage_2.1_ComfyUI/blob/f3d314f1e3f6f232bbe6c4088fbb5c8a7df26f0d/split_files/diffusion_models/hunyuanimage2.1_bf16.safetensors) | SHA256: 049634539c631a6c097d989efc2f29942637ddbb718a834a13d0c8cba22cf087 | No license is provided | no |
| 55 | qwen_image_edit_2509_fp8_e4m3fn.safetensors | 基于 20B 参数规模的 Qwen-Image-Edit 模型，继承其文字 / 文本渲染能力，并将其扩展至图像编辑任务，一致性优于基础模型 | unet | [Link](https://huggingface.co/Comfy-Org/Qwen-Image-Edit_ComfyUI/blob/87c96660003fddb262d73ed7539e823edecf59d4/split_files/diffusion_models/qwen_image_edit_2509_fp8_e4m3fn.safetensors) | SHA256: 318568f61951ab9da21100c7b896e3c1da67f0d2efad6421545e022cfaa2b2b4 | Apache-2.0 License | yes |
| 56 | lotus-depth-d-v1-1.safetensors | Lotus深度估计模型，用于生成高质量的深度图。在Qwen-Image InstantX ControlNet工作流中作为深度预处理组件，能够准确提取输入图像的空间深度信息，为3D感知的图像生成提供深度引导。 | unet | [Link](https://huggingface.co/Comfy-Org/lotus/resolve/main/lotus-depth-d-v1-1.safetensors) | dc1394219a04afdd9f72ad41ea0d3dfa435603fc7831a0ac5b884cc2b2ebf688 | Apache-2.0 License | yes |
| 57 | qwen_image_layered_fp8mixed.safetensors | Qwen-Image-Layered 主扩散模型的量化版本（fp8 mixed），来自 Comfy-Org/Qwen-Image-Layered_ComfyUI | unet | [Link](https://huggingface.co/Comfy-Org/Qwen-Image-Layered_ComfyUI/resolve/main/split_files/diffusion_models/qwen_image_layered_fp8mixed.safetensors) | SHA256: c9d51c2b144151d7eb12242567456ab7edda4f81f1a3b16c0b56e6bc59d9434c | Apache-2.0 License | yes |
| 58 | xinsir-all-in-one-sdxl.safetensors | ControlNet++ ProMax 模型，图像生成和编辑的一体化 | controlnet | [Link](https://huggingface.co/xinsir/controlnet-union-sdxl-1.0/blob/4d2716561322b77314e3bae98788af0a566ea766/diffusion_pytorch_model_promax.safetensors) | SHA256: 9fae2e50cb431bfcbe05822b59ec2228df545ef27f711dea8949e9f4ed9f7cdc | Apache-2.0 License | yes |
| 59 | control_depth-fp16.safetensors | ControlNet 的深度图生成模型 | controlnet | [Link](https://huggingface.co/webui/ControlNet-modules-safetensors/blob/5194dff6fe5e3d26310a12c527eae8bc02d3a482/control_depth-fp16.safetensors) | SHA256: f0a50531fa706fec36899d8499f228dcc917583b2de87e2f44f3374684b5c9c3 | No license is provided | no |
| 60 | flux-canny-controlnet-v3.safetensors | Black Forest Labs 的 FLUX.1-dev 模型的 Canny ControlNet  | controlnet | [Link](https://huggingface.co/XLabs-AI/flux-controlnet-canny-v3/blob/77521c5c0ccfa522984f3615bdc9f643c579824c/flux-canny-controlnet-v3.safetensors) | SHA256: 6546f29049796101a6370db0a43d2671d0294287c36b2b4e8792cf9e68f0eaf0 | FLUX.1 [dev] Non-Commercial License | no |
| 61 | flux-depth-controlnet-v3.safetensors | Black Forest Labs 的 FLUX.1-dev 模型的 Depth ControlNet  | controlnet | [Link](https://huggingface.co/XLabs-AI/flux-controlnet-depth-v3/blob/769482d36abd0eb453df91a8a20e72e48e1c2933/flux-depth-controlnet-v3.safetensors) | SHA256: d52eeaf8072de89d72b1ee79e3bdc79b5d795ed0a6881a029bbe13d833df7e5f | FLUX.1 [dev] Non-Commercial License | no |
| 62 | flux-hed-controlnet-v3.safetensors | Black Forest Labs 的 FLUX.1-dev 模型的 Hed ControlNet  | controlnet | [Link](https://huggingface.co/XLabs-AI/flux-controlnet-hed-v3/blob/d94a71e60daa97515597ce2eb0eced1fa9156420/flux-hed-controlnet-v3.safetensors) | SHA256: 31c110ae4557c19f1c2d74f824c20ea4c340161106df25cf7caf298e2f22f441 | FLUX.1 [dev] Non-Commercial License | no |
| 63 | control_v11p_sd15_openpose_fp16.safetensors | 图像引导生成模型, 控制人物动作和姿势, 社区提供的半精度模型格式 | controlnet | [Link](https://huggingface.co/comfyanonymous/ControlNet-v1-1_fp16_safetensors/blob/33142dc4c748128c527aa80a93e5f22a56a2c2ab/control_v11p_sd15_openpose_fp16.safetensors) | SHA256: 4003c1da17b0e4ba444e02140e1c0d83bb24b79e4dcfd613c3a554d38f0f89c7 | No license is provided | no |
| 64 | outfit2outfi-ControlNet.safetensors | 基于 ControlNet 的图像风格迁移工具，在保持主体、姿势和背景不变的情况下替换人物的穿着 | controlnet | [Link](https://civitai.com/models/191956/outfit-to-outfit-controlnet-outfit2outfit) | SHA256: 6AEC1DE323D72BA40F621D7A708C86F27E3E3A1FFE09336EE1E76F7284151BC2 | No license is provided | no |
| 65 | Qwen-Image-InstantX-ControlNet-Union.safetensors | Qwen-Image InstantX统一ControlNet模型，单个模型支持四种控制类型：Canny边缘检测、Soft Edge软边缘、Depth深度图和Pose姿态估计。专为Qwen-Image优化，提供精确的结构引导生成能力，可在ComfyUI中实现高质量的结构控制图像生成。 | controlnet | [Link](https://huggingface.co/Comfy-Org/Qwen-Image-InstantX-ControlNets/resolve/main/split_files/controlnet/Qwen-Image-InstantX-ControlNet-Union.safetensors) | d51dca0073366a675108d5b83c3b7ef941cf2214c9a1c95c23f1e9a228ddbdb0 | Apache-2.0 License | yes |
| 66 | diffusers_xl_depth_full.safetensors | SDXL controlnet 深度模型 | controlnet/XL | [Link](https://huggingface.co/diffusers/controlnet-depth-sdxl-1.0/blob/d409e4318358891d37974b9625c565e6cb8e9b2e/diffusion_pytorch_model.fp16.safetensors) | SHA256: 66a6813e6bd7270ecfe68206a59ddd605a011ae85321188376605c66e0a4f303 | CreativeML Open RAIL++-M License | yes |
| 67 | diffusion_pytorch_model.safetensors | InstantID官方提供的controlnet模型 | controlnet/instantid | [Link](https://huggingface.co/InstantX/InstantID/blob/0e9547646001e233e9ab3fde8625bf564edc2431/ControlNetModel/diffusion_pytorch_model.safetensors) | SHA256: c8127be9f174101ebdafee9964d856b49b634435cf6daa396d3f593cf0bbbb05 | Apache-2.0 License | yes |
| 68 | clip_g.safetensors | OpenCLIP bigG | clip | [Link](https://huggingface.co/lodestones/stable-diffusion-3-medium/blob/4a708bd3d18c10253247f8660cd4ffae6cd63bf1/stable-diffusion-3-medium/text_encoders/clip_g.safetensors) | SHA256: ec310df2af79c318e24d20511b601a591ca8cd4f1fce1d8dff822a356bcdb1f4 | stabilityai-ai-community | no |
| 69 | clip_l.safetensors | OpenAI CLIP-L | clip | [Link](https://huggingface.co/lodestones/stable-diffusion-3-medium/blob/4a708bd3d18c10253247f8660cd4ffae6cd63bf1/stable-diffusion-3-medium/text_encoders/clip_l.safetensors) | SHA256: 660c6f5b1abae9dc498ac2d21e1347d2abdb0cf6c0c0c8576cd796491d9a6cdd | stabilityai-ai-community | no |
| 70 | t5xxl_fp16.safetensors | Google T5-XXL | clip | [Link](https://huggingface.co/lodestones/stable-diffusion-3-medium/blob/4a708bd3d18c10253247f8660cd4ffae6cd63bf1/stable-diffusion-3-medium/text_encoders/t5xxl_fp16.safetensors) | SHA256: 6e480b09fae049a72d2a8c5fbccb8d3e92febeb233bbe9dfe7256958a9167635 | stabilityai-ai-community | no |
| 71 | t5xxl_fp8_e4m3fn.safetensors | comfyui官方提供的clip fp8 量化模型 | clip | [Link](https://huggingface.co/comfyanonymous/flux_text_encoders/blob/2f74b39c0606dae3b2196d79c18c2a40b71f3250/t5xxl_fp8_e4m3fn.safetensors) | SHA256: 7d330da4816157540d6bb7838bf63a0f02f573fc48ca4d8de34bb0cbfd514f09 | Apache-2.0 License | yes |
| 72 | t5xxl_fp8_e4m3fn_scaled.safetensors | FP8 量化版 T5 XXL 编码器权重文件，带缩放因子（scaled），用于高效低显存推理 | clip | [Link](https://huggingface.co/comfyanonymous/flux_text_encoders/blob/main/t5xxl_fp8_e4m3fn_scaled.safetensors) | SHA256: a498f0485dc9536735258018417c3fd7758dc3bccc0a645feaa472b34955557a | Apache-2.0 License | yes |
| 73 | qwen_2.5_vl_fp16.safetensors | 多模态语言模型 Qwen2.5-VL 作为文本编码器 | clip | [Link](https://huggingface.co/Comfy-Org/Omnigen2_ComfyUI_repackaged/blob/5c9d710be8de914e707c76dba1b1af5dd5a5aee9/split_files/text_encoders/qwen_2.5_vl_fp16.safetensors) | SHA256: ba05dd266ad6a6aa90f7b2936e4e775d801fb233540585b43933647f8bc4fbc3 | Apache-2.0 License | yes |
| 74 | umt5_xxl_fp8_e4m3fn_scaled.safetensors | Google 推出的超大规模多语言文本模型, UMT5-XXL 模型的 FP8（e4m3fn 格式）精度版本 | clip | [Link](https://huggingface.co/Comfy-Org/Wan_2.2_ComfyUI_Repackaged/blob/8b25e7c5754958ae8aac88fbdaa5cfb722f0666f/split_files/text_encoders/umt5_xxl_fp8_e4m3fn_scaled.safetensors) | SHA256: c3355d30191f1f066b26d93fba017ae9809dce6c627dda5f6a66eaa651204f68 | No license is provided | no |
| 75 | umt5_xxl_fp16.safetensors | Google 推出的超大规模多语言文本模型, UMT5-XXL 模型的 FP16 精度版本 | clip | [Link](https://huggingface.co/Comfy-Org/Wan_2.2_ComfyUI_Repackaged/blob/cd066aeeb9e2a1382bed53ed7e1fa2f1b55824d0/split_files/text_encoders/umt5_xxl_fp16.safetensors) | SHA256: 7b8850f1961e1cf8a77cca4c964a358d303f490833c6c087d0cff4b2f99db2af | No license is provided | no |
| 76 | qwen_2.5_vl_7b_fp8_scaled.safetensors | Qwen-Image 配套的文本编码器模型，支持多语言 Prompt 表达与图像编辑任务 | clip | [Link](https://huggingface.co/Comfy-Org/Qwen-Image_ComfyUI/blob/25608066f9bf5cdc28020836ce9549587053f346/split_files/text_encoders/qwen_2.5_vl_7b_fp8_scaled.safetensors) | SHA256: cb5636d852a0ea6a9075ab1bef496c0db7aef13c02350571e388aea959c5c0b4 | Apache-2.0 License | yes |
| 77 | qwen_2.5_vl_7b.safetensors | Qwen-Image 配套的文本编码器模型，支持多语言 Prompt 表达与图像编辑任务 | clip | [Link](https://huggingface.co/Comfy-Org/Qwen-Image_ComfyUI/blob/25608066f9bf5cdc28020836ce9549587053f346/split_files/text_encoders/qwen_2.5_vl_7b.safetensors) | SHA256: cfafd739459bc86257397259f612a9aee88e5b98e85b5c0d0d1717e898b3463a | Apache-2.0 License | yes |
| 78 | byt5_small_glyphxl_fp16.safetensors | Glyph-ByT5 small 文本编码器模型的权重 | clip | [Link](https://huggingface.co/Comfy-Org/HunyuanImage_2.1_ComfyUI/blob/1e58bcd2961da93ca5a196c8bda1c2039e324d2a/split_files/text_encoders/byt5_small_glyphxl_fp16.safetensors) | SHA256: 516910bb4c9b225370290e40585d1b0e6c8cd3583690f7eec2f7fb593990fb48 | No license is provided | no |
| 79 | t5_base.safetensors | T5-base 是 T5 模型的一种中等规模配置，适合在资源有限的环境中 | clip | [Link](https://huggingface.co/google-t5/t5-base/blob/fe6d9bf207cd3337512ca838a8b453f87a9178ef/model.safetensors) | a90903540cc02cbeb7ff9f823f1a80eb778c7e22426a0e620b01c77a5ec8f5b4 | Apache-2.0 License | yes |
| 80 | aesthetic.safetensors | Lora for FLUX,flux-lora工作流依赖的lora模型,https://civitai.com/models/633553 | loras | [Link](https://civitai.com/api/download/models/740450?type=Model&format=SafeTensor) | SHA256: 9c115fcd590efa77dbe57c29675730eeeade28c3dab10192fcc8008ed1ae65cc | FLUX.1 [dev] Non-Commercial License | no |
| 81 | ip-adapter-faceid-plusv2_sdxl_lora.safetensors | IP-Adapter-FaceID官方提供的lora模型 | loras | [Link](https://huggingface.co/h94/IP-Adapter-FaceID/blob/6eba56f2a87a7ebb2ef5a7737fe6c74df05f1f58/ip-adapter-faceid-plusv2_sdxl_lora.safetensors) | SHA256: f24b4bb2dad6638a09c00f151cde84991baf374409385bcbab53c1871a30cb7b | No license is provided | no |
| 82 | flux_realism_lora.safetensors | FLUX 的真实风格 LoRA | loras | [Link](https://huggingface.co/VideoAditor/Flux-Lora-Realism/blob/50072ae210ff40b5763517864be0bcc9aeeb5f46/flux_realism_lora.safetensors) | SHA256: 379e73dccfb57822ee3b12f374e141ce1c79a13b7ff19da4219ef2a2a610038e | No license is provided | no |
| 83 | Hoko-v0.9.safetensors | Hoko 漫画风格的 LoRA 模型，来自 https://civitai.com/models/647408?modelVersionId=724312 | loras | [Link](https://civitai.com/api/download/models/724312?type=Model&format=SafeTensor) | SHA256: 711080B80A155999A73FFC051EBC443597B876DA9E8DE57DD29D54BD20289717 | No license is provided | no |
| 84 | alimama-creative/LUX.1-Turbo-Alpha.safetensors | 阿里妈妈团队发布的基于FLUX.1-dev模型的8步蒸馏版 | loras | [Link](https://huggingface.co/alimama-creative/FLUX.1-Turbo-Alpha/blob/dac4ce8aa51dff7ff47ce4a32d1a44466126237d/diffusion_pytorch_model.safetensors) | SHA256: 77f7523a5e9c3da6cfc730c6b07461129fa52997ea06168e9ed5312228aa0bff | FLUX.1 [dev] Non-Commercial License | no |
| 85 | ghibliStyle_kontext_byJaneB.safetensors | 专为 Flux Kontext 工作流设计的 Ghibli 风格 LoRA 权重，生成吉卜力动画风格图像 | loras | [Link](https://civitai.com/models/1732367/ghibli-style-flux-kontext) | SHA256: 7A711E6C2A1C25E869F98E168AB99E236D8ABC1CEA710994A6FE599F8CA8BA28 | FLUX.1 [dev] Non-Commercial License | no |
| 86 | kontext_change_clothes_t1.safetensors | 基于 FLUX.1 Kontext 的图像编辑 LoRA，专门用于通过文字提示无缝替换人物的服装样式 | loras | [Link](https://civitai.com/models/1729159/kontextchangeclothes) | 8FF1CB258F6542A0A31138C7F544E05EB176BE2CB280A1D4245B801DD48FF961 | FLUX.1 [dev] Non-Commercial License | no |
| 87 | Qwen-Image-Lightning-4steps-V1.0.safetensors | Qwen-Image-Lightning-4steps-V1.0 是基于 Qwen-Image 的蒸馏式 LoRA，支持仅 4 步推理，速度提升约 12–25 倍，且画质在多数场景基本保持不变 | loras | [Link](https://huggingface.co/lightx2v/Qwen-Image-Lightning/blob/main/Qwen-Image-Lightning-4steps-V1.0.safetensors) | SHA256: 9526e90d71c4290392feeccf3c2172cb77ab3a489f1faeb956637f97acb4c8b1 | Apache-2.0 License | yes |
| 88 | lightx2v_I2V_14B_480p_cfg_step_distill_rank64_bf16.safetensors | Lightx2v 团队的 I2V 14B 480p 蒸馏 LoRA（rank64，Cfg+Step Distill），用于 Wan2.x 系列的图像转视频加速与质量提升 | loras | [Link](https://huggingface.co/Kijai/WanVideo_comfy/blob/main/Lightx2v/lightx2v_I2V_14B_480p_cfg_step_distill_rank64_bf16.safetensors) | SHA256: 85c4a61c30e0497aa44b91d93a893b624708461a56fe5485183b28fa07e2dfb3 | Apache-2.0 License | yes |
| 89 | WanAnimate_relight_lora_fp16.safetensors | Wan2.2 Animate 专用 Relight LoRA，用于统一角色与场景的打光与光照融合，配合 Animate 工作流提升灯光一致性 | loras | [Link](https://huggingface.co/Kijai/WanVideo_comfy/blob/main/LoRAs/Wan22_relight/WanAnimate_relight_lora_fp16.safetensors) | SHA256: fc646c74c73f4b251f5fd9bc440ef21b03b27305f499966c68b2b3aa31498561 | No license is provided | no |
| 90 | wan2.2_i2v_lightx2v_4steps_lora_v1_low_noise.safetensors | Lightx2v LoRA，用于图像到视频任务的低噪声版本，增强视频稳定性 | loras | [Link](https://huggingface.co/Comfy-Org/Wan_2.2_ComfyUI_Repackaged/resolve/main/split_files/loras/wan2.2_i2v_lightx2v_4steps_lora_v1_low_noise.safetensors) | SHA256: 56f7f014ff2fa7a78d5ad570a73480f130c1b6161efc7a845f379b894a505c7e | No license is provided | no |
| 91 | wan2.2_i2v_lightx2v_4steps_lora_v1_high_noise.safetensors | Lightx2v LoRA，用于图像到视频任务的高噪声版本，增强视频稳定性 | loras | [Link](https://huggingface.co/Comfy-Org/Wan_2.2_ComfyUI_Repackaged/resolve/main/split_files/loras/wan2.2_i2v_lightx2v_4steps_lora_v1_high_noise.safetensors) | SHA256: 6c7e0a5e7a69788a1e13b6d7e4be357c6bfe6a839315f55f1a4c8a6b7895974e | No license is provided | no |
| 92 | OpenCLIP-ViT-H-14-laion2B-s32B-b79K.safetensors | 基于 OpenCLIP 通过 LAION-2B 训练的 CLIP ViT-H/14 模型。 | clip_vision | [Link](https://huggingface.co/laion/CLIP-ViT-H-14-laion2B-s32B-b79K/blob/1c2b8495b28150b8a4922ee1c8edee224c284c0c/open_clip_model.safetensors) | SHA256: 0084e75319a50ad85ef45377bad5bc38f2f58824459eb690048d51c9f8863be5 | MIT License | yes |
| 93 | CLIP-ViT-H-14-laion2B-s32B-b79K.safetensors | IP-Adapter官方提供的clip-vision模型 | clip_vision | [Link](https://huggingface.co/h94/IP-Adapter/blob/0859e809306db97aa2338e370587ab284e8a754f/models/image_encoder/model.safetensors) | SHA256: 6ca9667da1ca9e0b0f75e46bb030f7e011f44f86cbfb8d5a36590fcd7507b030 | Apache-2.0 License | yes |
| 94 | sigclip_vision_patch14_384.safetensors | Flux Redux 工作流提供的专用 CLIP Vision 编码器模型，基于 Google 的 SigLIP SoViT‑400M‑Patch14‑384 架构，负责从参考图提取图像特征以用于风格控制和图像迁移 | clip_vision | [Link](https://huggingface.co/Comfy-Org/sigclip_vision_384/blob/f44ecf2517296ea3d06aa1e9b6c6446beda90094/sigclip_vision_patch14_384.safetensors) | SHA256: 1fee501deabac72f0ed17610307d7131e3e9d1e838d0363aa3c2b97a6e03fb33 | No license is provided | no |
| 95 | clip_vision_h.safetensors | Wan2.1/Wan Animate 工作流配套的 CLIP Vision H 模型（ViT-H/14 变体），用于从参考帧/图像中提取视觉特征 | clip_vision | [Link](https://huggingface.co/Comfy-Org/Wan_2.1_ComfyUI_repackaged/blob/main/split_files/clip_vision/clip_vision_h.safetensors) | SHA256: 64a7ef761bfccbadbaa3da77366aac4185a6c58fa5de5f589b42a65bcc21f161 | Apache-2.0 License | yes |
| 96 | depth_anything_vitl14.pth | Depth Anything 单目深度估计模型 | comfyui_controlnet_aux/ckpts/LiheYoung/Depth-Anything/checkpoints | [Link](https://huggingface.co/spaces/LiheYoung/Depth-Anything/blob/843bd9726dbd81cae08889eecfbf242e33b6b35b/checkpoints/depth_anything_vitl14.pth) | SHA256: 6c6a383e33e51c5fdfbf31e7ebcda943973a9e6a1cbef1564afe58d7f2e8fe63 | No license is provided | no |
| 97 | depth_anything_v2_vitl.pth | Depth Anything V2 单目深度估计模型 | comfyui_controlnet_aux/ckpts/depth-anything/Depth-Anything-V2-Large | [Link](https://huggingface.co/depth-anything/Depth-Anything-V2-Large/blob/c88f92d49d0b635fb6c088199cce0c969807dccc/depth_anything_v2_vitl.pth) | SHA256: a7ea19fa0ed99244e67b624c72b8580b7e9553043245905be58796a608eb9345 | CC-BY-NC-4.0 License | no |
| 98 | ControlNetHED.pth | HED权重，soft HED边界将保留输入图像中的许多细节，适合重新着色和风格化 | comfyui_controlnet_aux/ckpts/lllyasviel/Annotators | [Link](https://huggingface.co/lllyasviel/Annotators/blob/9cf9fa959ab9a81454b68a082c57052e671677e6/ControlNetHED.pth) | SHA256: 5ca93762ffd68a29fee1af9d495bf6aab80ae86f08905fb35472a083a4c7a8fa | Apache-2.0 License | yes |
| 99 | svdq-int4_r32-flux.1-kontext-dev.safetensors | flux.1-kontext 的 nunchaku int4 模型，兼容性更佳，FP4 仅支持 RTX 50 系列。 | diffusion_models | [Link](https://huggingface.co/mit-han-lab/nunchaku-flux.1-kontext-dev/blob/cff45ce4e9ddee4bd8920f7df53f835dd4c32191/svdq-int4_r32-flux.1-kontext-dev.safetensors) | SHA256: 5dceaae1666bd76402bb9e47a9431a51b2b528f921502ee92b90257717acbebb | FLUX.1 [dev] Non-Commercial License | no |
| 100 | Wan2_2-Animate-14B_fp8_e4m3fn_scaled_KJ.safetensors | Wan2.2 Animate 14B 的 FP8（e4m3fn_scaled）量化版 Diffusion 模型，适合低显存高效生成角色动画与视频 | diffusion_models | [Link](https://huggingface.co/Kijai/WanVideo_comfy_fp8_scaled/blob/main/Wan22Animate/Wan2_2-Animate-14B_fp8_e4m3fn_scaled_KJ.safetensors) | SHA256: 2936b31473a967e7a429a6646bba60e7862d0938e178b58b2a140f391dd5b8e6 | Apache-2.0 License | yes |
| 101 | wan2.2_animate_14B_bf16.safetensors | Wan2.2 Animate 14B 的 bf16 精度文本/图像生成视频模型，Comfy-Org 官方 Wan2.2_ComfyUI 重打包版本 | diffusion_models | [Link](https://huggingface.co/Comfy-Org/Wan_2.2_ComfyUI_Repackaged/blob/main/split_files/diffusion_models/wan2.2_animate_14B_bf16.safetensors) | SHA256: 7d37cb0120488dec2a061135aa3018426925318a60838c3de935c829abc88667 | Apache-2.0 License | yes |
| 102 | wan2.2_s2v_14B_fp8_scaled.safetensors | Wan2.2 图像到视频模型的 FP8 量化版本，适合高效生成动画视频 | diffusion_models | [Link](https://huggingface.co/Comfy-Org/Wan_2.2_ComfyUI_Repackaged/resolve/main/split_files/diffusion_models/wan2.2_s2v_14B_fp8_scaled.safetensors) | SHA256: 618fa7a9e7d3147c6705905f3c0ac607e86e4dbed271be59c8e6bb0327f8575e | Apache-2.0 License | yes |
| 103 | hunyuanvideo1.5_720p_t2v_fp16.safetensors | HunyuanVideo 1.5 720p 文本转视频主模型，FP16 精度，适用于 ComfyUI 文本生成视频工作流 | diffusion_models | [Link](https://huggingface.co/Comfy-Org/HunyuanVideo_1.5_repackaged/resolve/main/split_files/diffusion_models/hunyuanvideo1.5_720p_t2v_fp16.safetensors) | SHA256: 4f05e2a31b51b7ddee3ef336d234e0932001cbdc85887da15451069fda3e1d05 | tencent-hunyuan-community | N/A |
| 104 | hunyuanvideo1.5_720p_i2v_fp16.safetensors | HunyuanVideo 1.5 图像转视频（Image-to-Video）720p 主模型，FP16 精度，适用于 I2V 工作流 | diffusion_models | [Link](https://huggingface.co/Comfy-Org/HunyuanVideo_1.5_repackaged/resolve/main/split_files/diffusion_models/hunyuanvideo1.5_720p_i2v_fp16.safetensors) | No hash is provided | tencent-hunyuan-community | N/A |
| 105 | flux2_dev_fp8mixed.safetensors | FLUX.2 [dev] 主扩散模型的量化版本（fp8 mixed），来自 Comfy-Org/flux2-dev，用于文本/图像生成与编辑 | diffusion_models | [Link](https://huggingface.co/Comfy-Org/flux2-dev/resolve/main/split_files/diffusion_models/flux2_dev_fp8mixed.safetensors) | No hash is provided | FLUX.2 [dev] Non-Commercial License | N/A |
| 106 | efficient_sam_s_gpu.jit | 运行ComfyUI-YoloWorld-EfficientSAM所依赖的权重 | esam | [Link](https://huggingface.co/spaces/SkalskiP/YOLO-World/blob/7500bfd7f0e722269e158889b28800c9ddccb4cc/efficient_sam_s_gpu.jit) | SHA256: e47c589ead2c6a80d38050ce63083a551e288db27113d534e0278270fc7cba26 | Apache-2.0 License | yes |
| 107 | efficient_sam_s_cpu.jit | 运行ComfyUI-YoloWorld-EfficientSAM所依赖的权重 | esam | [Link](https://huggingface.co/spaces/SkalskiP/YOLO-World/blob/7500bfd7f0e722269e158889b28800c9ddccb4cc/efficient_sam_s_cpu.jit) | SHA256: 8b63ab268e9020b0fb7fc9f46e742644d4c9ea6e5d9caf56045f0afb6475db09 | Apache-2.0 License | yes |
| 108 | EVA02_CLIP_L_336_psz14_s6B.pt | 高精度的 ViT-L/14 文本-图像嵌入模型 | eva_clip | [Link](https://huggingface.co/QuanSun/EVA-CLIP/blob/690a5975ca06e9a43d16bae440dd08213bfff0c5/EVA02_CLIP_L_336_psz14_s6B.pt) | SHA256: 84c3a17a228c567a155259b2245b0b59072bf7da510260a0a02ec54de6d50b05 | MIT License | yes |
| 109 | parsing_bisenet.pth | Facexlib所需权重, 用于PulidFlux的面部五官区域分割 | facexlib | [Link](https://github.com/xinntao/facexlib/releases/download/v0.2.0/parsing_bisenet.pth) | SHA256: 468e13ca13a9b43cc0881a9f99083a430e9c0a38abd935431d1c28ee94b26567 | MIT License | yes |
| 110 | parsing_parsenet.pth | Facexlib所需权重, 用于PulidFlux的面部区域分割 | facexlib | [Link](https://github.com/xinntao/facexlib/releases/download/v0.2.2/parsing_parsenet.pth) | SHA256: 3d558d8d0e42c20224f13cf5a29c79eba2d59913419f945545d8cf7b72920de2 | MIT License | yes |
| 111 | detection_Resnet50_Final.pth | Facexlib所需权重, 用于PulidFlux的人脸检测 | facexlib | [Link](https://github.com/xinntao/facexlib/releases/download/v0.1.0/detection_Resnet50_Final.pth) | SHA256: 6d1de9c2944f2ccddca5f5e010ea5ae64a39845a86311af6fdf30841b0a5a16d | MIT License | yes |
| 112 | groundingdino_swinb_cogcoor.pth | GroundingDINO SwinB 模型，用于开放词汇的目标检测 | groundingdino | [Link](https://huggingface.co/ShilongLiu/GroundingDINO/blob/81ca292d6957b64598c0927d23f9ca399900e170/groundingdino_swinb_cogcoor.pth) | SHA256: 46270f7a822e6906b655b729c90613e48929d0f2bb8b9b76fd10a856f3ac6ab7 | Apache-2.0 License | yes |
| 112 | models/antelopev2 | Insightface官方权重, 包含1k3d68.onnx, 2d106det.onnx, genderage.onnx, glintr100.onnx, scrfd_10g_bnkps.onnx模型 | insightface | [Link](https://github.com/deepinsight/insightface/releases/download/v0.7/antelopev2.zip) | No hash is provided | Under MIT License (non-commercial) | no |
| 113 | models/buffalo_l | Insightface官方权重, 包含1k3d68.onnx, 2d106det.onnx, det_10g.onnx, genderage.onnx, w600k_r50.onnx模型 | insightface | [Link](https://github.com/deepinsight/insightface/releases/download/v0.7/buffalo_l.zip) | No hash is provided | Under MIT License (non-commercial) | no |
| 114 | ip-adapter.bin | InstantID官方提供的主干模型 | instantid | [Link](https://huggingface.co/InstantX/InstantID/blob/e017855ddd75ce2f720d0198647d0c686830202a/ip-adapter.bin) | SHA256: 02b3618e36d803784166660520098089a81388e61a93ef8002aa79a5b1c546e1 | Apache-2.0 License | yes |
| 115 | ip-adapter-plus_sdxl_vit-h.safetensors | SDXL 的 IP-Adapter 模型，适用于图像风格迁移 | ipadapter | [Link](https://huggingface.co/h94/IP-Adapter/blob/9bf28b38530e55ffa91c6d82e5161a982c22f284/sdxl_models/ip-adapter-plus_sdxl_vit-h.safetensors) | SHA256: 3f5062b8400c94b7159665b21ba5c62acdcd7682262743d7f2aefedef00e6581 | Apache-2.0 License | yes |
| 116 | ip-adapter-faceid-plusv2_sdxl.bin | IP-Adapter-FaceID官方提供的主干模型 | ipadapter | [Link](https://huggingface.co/h94/IP-Adapter-FaceID/blob/36ce7f96f0c76c1e8b100f10fa094a2ac48ea1c6/ip-adapter-faceid-plusv2_sdxl.bin) | SHA256: c56e7f9dce0ab4aae955e614d2e172f2d51ddf95e3f7983469504f59b3e31c32 | No license is provided | no |
| 117 | ip-adapter-plus_sd15.safetensors | 基于 Stable Diffusion 1.5 的 IP‑Adapter Plus 版本，运用更细致的图像提示，以增强图像生成效果，兼容 SD1.5 IP‑Adapter 编码器 | ipadapter | [Link](https://huggingface.co/h94/IP-Adapter/blob/9bf28b38530e55ffa91c6d82e5161a982c22f284/models/ip-adapter-plus_sd15.safetensors) | SHA256: a1c250be40455cc61a43da1201ec3f1edaea71214865fb47f57927e06cbe4996 | Apache-2.0 License | yes |
| 118 | clip-vit-h-14-laion2B-s32B-b79K.safetensors | 基于 OpenCLIP 的 ViT‑H/14 模型，支持高性能的零样本图像分类与图文检索应用 | ipadapter | [Link](https://huggingface.co/laion/CLIP-ViT-H-14-laion2B-s32B-b79K/blob/main/open_clip_model.safetensors) | SHA256: 0084e75319a50ad85ef45377bad5bc38f2f58824459eb690048d51c9f8863be5 | MIT License | yes |
| 119 | landmark.onnx | LivePortrait所需的面部地标检测模型, 从驱动视频帧中提取面部关键点, 使真实的运动转移到静态图像 | liveportrait | [Link](https://huggingface.co/Kijai/LivePortrait_safetensors/blob/9e2ebb628f4c5046a54d7cbfe044edb124e8dc0a/landmark.onnx) | SHA256: 31d22a5041326c31f19b78886939a634a5aedcaa5ab8b9b951a1167595d147db | No license is provided | no |
| 120 | appearance_feature_extractor.safetensors | LivePortrait 所需的外观特征提取模型, 从静态图像中提取人物的外观特征（如肤色、纹理、五官风格等） | liveportrait | [Link](https://huggingface.co/Kijai/LivePortrait_safetensors/blob/9e2ebb628f4c5046a54d7cbfe044edb124e8dc0a/appearance_feature_extractor.safetensors) | SHA256: 38bef5de50a92bf1fc66e8c511051a19dfacdf80c37f8713425ec15dc9ca7d34 | No license is provided | no |
| 121 | motion_extractor.safetensors | LivePortrait 所需的运动特征提取模型， 从驱动视频中提取面部动作与表情动态（如嘴型、眼神、头部姿态等） | liveportrait | [Link](https://huggingface.co/Kijai/LivePortrait_safetensors/blob/9e2ebb628f4c5046a54d7cbfe044edb124e8dc0a/motion_extractor.safetensors) | SHA256: 3568cd410e29d046771acb55ecfdfe4c7c197d345bd8b7f95942ef63130b6c9e | No license is provided | no |
| 122 | warping_module.safetensors | LivePortrait 所需的图像变形模块, 根据提取到的运动特征与人脸关键点，对静态图像进行空间变换 | liveportrait | [Link](https://huggingface.co/Kijai/LivePortrait_safetensors/blob/9e2ebb628f4c5046a54d7cbfe044edb124e8dc0a/warping_module.safetensors) | SHA256: f7b7834bd6039b4088f72e5161e60ad366f68a3763df8a3eac0bc0f9d46fdbbf | No license is provided | no |
| 123 | spade_generator.safetensors | LivePortrait 所需的图像生成模块, 接收来自身份特征、表情动作与变形模块的中间特征, 生成最终动态图像 | liveportrait | [Link](https://huggingface.co/Kijai/LivePortrait_safetensors/blob/9e2ebb628f4c5046a54d7cbfe044edb124e8dc0a/spade_generator.safetensors) | SHA256: ca04fbec765745e9eae836d2d7522c274647b277ce5f25104fa1705b75222212 | No license is provided | no |
| 124 | stitching_retargeting_module.safetensors | LivePortrait 系统中的图像拼接与重定向模块, 将生成的人脸区域与原始背景或上下文图像进行融合 | liveportrait | [Link](https://huggingface.co/Kijai/LivePortrait_safetensors/blob/9e2ebb628f4c5046a54d7cbfe044edb124e8dc0a/stitching_retargeting_module.safetensors) | SHA256: 60725cf3523ae413880da28ed583c9e84c5c25695a0cef1b77210ea31cc424ea | No license is provided | no |
| 125 | wan2.2_ti2v_5B_fp16.safetensors | Wan2.2 模型中支持文本到视频和图像到视频的混合版本，50亿参数，单个模型满足两个核心任务需求的 Diffusion 模型权重文件，采用 FP16 精度格式 | model/diffusion_models | [Link](https://huggingface.co/Comfy-Org/Wan_2.2_ComfyUI_Repackaged/blob/5ca2dfecf59320b1d4605b5802e64f77a8676afe/split_files/diffusion_models/wan2.2_ti2v_5B_fp16.safetensors) | SHA256: 456f901338bd9eadbded3828b819109a9b68e8a525ca5cf8d0049a69fcfeca1e | Apache-2.0 License | yes |
| 126 | pulid_flux_v0.9.0.safetensors | PulidFlux 推理专用模型, 融合了 Flux 语义特征, 优化小显存环境下的 Diffusion 生成 | pulid | [Link](https://huggingface.co/guozinan/PuLID/blob/98602ea62f42e26ccfce3f50fd409b0d4adf531c/pulid_flux_v0.9.0.safetensors) | SHA256: 14e3c257ae88643412db0f6d239f55132cd5487107ed33944c54339284cbe1e6 | Apache-2.0 License | yes |
| 127 | RMBG-1.4.pth | 基于 RMBG 1.4 版本的背景去除模型，实现抠图，提取前景 | rembg | [Link](https://huggingface.co/briaai/RMBG-1.4/blob/05299b644683d6a74c5c2204baa63304c0bee516/model.pth) | SHA256: 893c16c340b1ddafc93e78457a4d94190da9b7179149f8574284c83caebf5e8c | bria-rmbg-1.4 License | no |
| 128 | sam_vit_b_01ec64.pth | ComfyUI-Impact-Pack官方提供的sam模型 | sam | [Link](https://huggingface.co/licyk/comfyui-extension-models/blob/bf831f04b23a9337df320d0522f0ca3d2c76cdd8/ComfyUI-Impact-Pack/sam_vit_b_01ec64.pth) | SHA256: ec2df62732614e57411cdcf32a23ffdf28910380d03139ee0f4fcbe91eb8c912 | CreativeML Open RAIL -M License | yes |
| 129 | sam2.1_hiera_large.pt | Segment Anything 模型的改进版本，实现更高精度的图像任意区域分割 | sam2 | [Link](https://huggingface.co/facebook/sam2.1-hiera-large/blob/227a114a2f535cd147f82442e7d2038cdd2e5d68/sam2.1_hiera_large.pt) | SHA256: 2647878d5dfa5098f2f8649825738a9345572bae2d4350a2468587ece47dd318 | Apache-2.0 License | yes |
| 130 | flux1-redux-dev.safetensors | 专为 FLUX.1 系列模型（flux dev 或 flux schnell）设计的轻量级风格适配器模型，允许输入一张或多张参考图像，从中提取风格并生成一致风格的图像变体，支持结合文本重新塑造图 | style_models | [Link](https://huggingface.co/black-forest-labs/FLUX.1-Redux-dev/blob/a1b3bdcb4bdc58ce04874b9ca776d61fc3e914bb6beab41efb63e4e2694dca45/flux1-redux-dev.safetensors) | SHA256: a1b3bdcb4bdc58ce04874b9ca776d61fc3e914bb6beab41efb63e4e2694dca45 | FLUX.1 [dev] Non-Commercial License | no |
| 131 | gemma_2_2b_fp16.safetensors | 基于 Gemma 2（2B 参数规模）的 FP16 精度文本编码器权重，生成高质量的文本条件编码 | text_encoders | [Link](https://huggingface.co/neta-art/Neta-Lumina/blob/97dd97806a9bb7c2cbc1f63ba77bb206831de9e7/Text%20Encoder/gemma_2_2b_fp16.safetensors) | SHA256: 29761442862f8d064d3f854bb6fabf4379dcff511a7f6ba9405a00bd0f7e2dbd | Apache-2.0 License | yes |
| 132 | mistral_3_small_flux2_fp8.safetensors | 基于 Mistral Small 3 的 FLUX.2 文本编码器，FP8 量化版本，适合低显存推理使用 | text_encoders | [Link](https://huggingface.co/Comfy-Org/flux2-dev/resolve/main/split_files/text_encoders/mistral_3_small_flux2_fp8.safetensors) | No hash is provided | Apache-2.0 License | yes |
| 133 | roboflow/clip/ViT-B-32/textual.onnx | roboflow api 下载的模型，没有永久有效的下载地址，只能通过代码下载。`from inference.models import Clip; Clip(model_id='clip/ViT-B-32')` | third_party/inference/clip/ViT-B-32 | [Link](None) | SHA256: e06418e5b77eda0bf7689656fe4f2759bf6a8fee6cea3566e43536cda860e8d8 | AGPL-3.0 License | yes |
| 134 | roboflow/clip/ViT-B-32/visual.onnx | roboflow api 下载的模型，没有永久有效的下载地址，只能通过代码下载。`from inference.models import Clip; Clip(model_id='clip/ViT-B-32')` | third_party/inference/clip/ViT-B-32 | [Link](None) | SHA256: 1c5a1ca6ec0ee440b47a21cdb9183f9c3491ab203ed62aa68c02340b91eb58f8 | AGPL-3.0 License | yes |
| 135 | roboflow/yolo_world/l/yolo-world.pt | roboflow api 下载的模型，没有永久有效的下载地址，只能通过代码下载。`from inference.models import YOLOWorld; YOLOWorld(model_id='yolo_world/l')` | third_party/inference/yolo_world/l | [Link](None) | SHA256: dab1b559400c8c6786757ce48cb5a83ca0f21550e8ed9ef6af6cb0e5a0c14845 | AGPL-3.0 License | yes |
| 136 | bbox/face_yolov8m.pt | YOLOv8 权重，用于目标检测 | ultralytics | [Link](https://huggingface.co/datasets/Gourieff/ReActor/blob/a99aa96d1d9b2896b9ec9843ae01da4ad28cddf3/models/detection/bbox/face_yolov8m.pt) | SHA256: e3893a92c5c1907136b6cc75404094db767c1e0cfefe1b43e87dad72af2e4c9f | No license is provided | no |
| 137 | 4x_foolhardy_Remacri.pth | 用于图像放大的4x超分模型，Remacri风格 | upscale_models | [Link](https://huggingface.co/FacehugmanIII/4x_foolhardy_Remacri/blob/8caa4f487df740945da9949a68a6ad7836074ece/4x_foolhardy_Remacri.pth) | SHA256: e1a73bd89c2da1ae494774746398689048b5a892bd9653e146713f9df8bca86a | No license is provided | no |
| 138 | 4xUltrasharp_4xUltrasharpV10.pt | 图像细节增强模型 | upscale_models | [Link](https://civitai.com/api/download/models/125843?type=Model&format=PickleTensor) | SHA256: A5812231FC936B42AF08A5EDBA784195495D303D5B3248C24489EF0C4021FE01 | Attribution-NonCommercial-ShareAlike 4.0 International | no |
| 139 | wd-v1-4-moat-tagger-v2.csv | 标签词列表, 供 wd-v1-4-moat-tagger 使用 | wd14_tagger | [Link](https://huggingface.co/SmilingWolf/wd-v1-4-moat-tagger-v2/blob/8452cddf280b952281b6e102411c50e981cb2908/selected_tags.csv) | SHA256: 8c8750600db36233a1b274ac88bd46289e588b338218c2e4c62bbc9f2b516368 | Apache-2.0 License | yes |
| 140 | wd-v1-4-moat-tagger-v2.onnx | wd-v1-4-moat-tagger 本体模型, 进行图像标签自动标注 | wd14_tagger | [Link](https://huggingface.co/SmilingWolf/wd-v1-4-moat-tagger-v2/blob/8452cddf280b952281b6e102411c50e981cb2908/model.onnx) | SHA256: b8cef913be4c9e8d93f9f903e74271416502ce0b4b04df0ff1e2f00df488aa03 | Apache-2.0 License | yes |

> 说明： `ComfyUIPath` 一栏表示该模型文件在 `siliconflow/awesome-comfyui-models/<ComfyUIPath>` 对应文件夹里。
启动 ComfyUI 时，模型文件要被放到 `ComfyUI/models/<ComfyUIPath>` 对应文件夹里。
WeightID 一栏用于唯一标识每个模型权重文件。

### 模型下载
> 您可以通过如下git clone命令，或者ModelScope SDK来下载模型

SDK下载
```bash
#安装ModelScope
pip install modelscope
```
```python
#SDK模型下载
from modelscope import snapshot_download
model_dir = snapshot_download('siliconflow/awesome-comfyui-models')
```

Git下载
```
#Git模型下载
git clone https://www.modelscope.cn/siliconflow/awesome-comfyui-models.git
```

<p style="color: lightgrey;">如果您是本模型的贡献者，我们邀请您根据<a href="https://modelscope.cn/docs/ModelScope%E6%A8%A1%E5%9E%8B%E6%8E%A5%E5%85%A5%E6%B5%81%E7%A8%8B%E6%A6%82%E8%A7%88" style="color: lightgrey; text-decoration: underline;">模型贡献文档</a>，及时完善模型卡片内容。</p>