comfyanonymous
🌟 LongCat-Image-Edit
Feature Idea
We introduce LongCat-Image-Edit, the image editing version of Longcat-Image. LongCat-Image-Edit supports bilingual (Chinese-English) editing, achieves state-of-the-art performance among open-source image editing models, delivering leading instruction-following and image quality with superior visual consistency.
[LongCat-Image](https://huggingface.co/meituan-longcat/LongCat-Image)
https://huggingface.co/meituan-longcat/LongCat-Image-Edit
Key Features 主要特点 🌟 Superior Precise Editing: LongCat-Image-Edit supports various editing tasks, such as global editing, local editing, text modification, and reference-guided editing. It has strong semantic understanding capabilities and can perform precise editing according to instructions. 🌟 超强精准编辑 ：LongCat-Image-Edit 支持多种编辑任务，例如全局编辑、局部编辑、文本修改和参考引导编辑。它拥有强大的语义理解能力，能够根据指令进行精准编辑。 🌟 Consistency Preservation: LongCat-Image-Edit has strong consistency preservation capabilities, specifically scrutinizes whether attributes in non-edited regions, such as layout, texture, color tone, and subject identity, remain invariant unless targeted by the instruction, is well demonstrated in multi-turn editing. 🌟 一致性保持 ：LongCat-Image-Edit 具有强大的一致性保持能力，特别是会仔细检查未编辑区域中的属性（如布局、纹理、色调和主题标识）是否保持不变，除非受到指令的指示，这在多轮编辑中得到了很好的体现。 🌟 Strong Benchmark Performance: LongCat-Image-Edit achieves state-of-the-art (SOTA) performance in image editing tasks while significantly improving model inference efficiency, especially among open-source image editing models. 🌟 强大的基准性能 ：LongCat-Image-Edit 在图像编辑任务中达到了最先进的 (SOTA) 性能，同时显著提高了模型推理效率，尤其是在开源图像编辑模型中。
Existing Solutions
No response
Other
No response
可能性极大，comfyui对图像模型的支持力度非常大，绝大多数都支持了。cat 比ovi要好
ovis使用起来感觉没z-image稳定，性能z-image也要好些。
https://github.com/meituan-longcat/LongCat-Image/issues/8
ovis 这种就是kpi ocr模型，没有价值
这有一个非官方实现，看起来不错。
试了效果还是非常不错的，就是好像不支持多图，要是能保持多参考图人物的一致性将秒杀qwen edit
I believe the model might be possible to take in multiple image inputs, because official pipeline implementation allows passing a list of images as input.
In china meituan is almost as famous and big as alibaba, we need this support.Qwen edit has almost a half unused layers, too big
I was waiting for ComfyUI to support this natively, but it seems there’s very little interest. I already created a ComfyUI node for it, but it’s not production-ready because I assumed native support was coming, so I didn’t work on it, beyond testing. But looks like I should test it properly and release it.
I even quantized it and created GGUF version.
https://huggingface.co/vantagewithai/LongCat-Image-GGUF https://huggingface.co/vantagewithai/LongCat-Image-Edit-GGUF
I successfully run LongCat-Image/LongCat-Image-Edit using the node https://github.com/sooxt98/comfyui_longcat_image, and you can experience it online on the cnb platform.
https://cnb.cool/fuyaotingyu/ComfyUI/ComfyUI-LongCat-Image https://cnb.cool/fuyaotingyu/ComfyUI/ComfyUI-LongCat-Image-Edit
If you want to run it locally, please refer to the list of commands below.
cd /workspace/ComfyUI/custom_nodes
git clone https://github.com/sooxt98/comfyui_longcat_image.git
cd comfyui_longcat_image
pip install -r requirements.txt --no-cache-dir
pip install git+https://github.com/meituan-longcat/LongCat-Image.git
You also need to confirm that you have downloaded the complete model weight file. The command to download it from huggingface is as follows:
pip install "huggingface_hub[cli]"
# For text-to-image
huggingface-cli download meituan-longcat/LongCat-Image --local-dir /workspace/ComfyUI/models/diffusion_models/LongCat-Image
# For image editing
huggingface-cli download meituan-longcat/LongCat-Image-Edit --local-dir /workspace/ComfyUI/models/diffusion_models/LongCat-Image-Edit
Note: You should replace the root path of your local ComfyUI.
@llnancy The custom node you posted simply wraps around the official inference code, which itself simply inherits from the diffusers library.
What we want is native integration of LongCat Image and the editing model into ComfyUI, not a naive wrapper function that is incompatible with the rest of the core ComfyUI nodes.
@llnancy The custom node you posted simply wraps around the official inference code, which itself simply inherits from the
diffuserslibrary.What we want is native integration of LongCat Image and the editing model into ComfyUI, not a naive wrapper function that is incompatible with the rest of the core ComfyUI nodes.
I understand what you mean, and that's what I expect as well. I'm just providing an available method before official support is available.
z image base and edit is coming ,maybe longcat will never be supported. Such a pity, 6B powerful edit model, z image edit most likely >10B. @comfyanonymous
