Image editing with AI has come a long way, but it's still frustrating in many real world cases. You give a simple instruction, and the output looks almost right but not quite. Faces change between edits, characters lose consistency, group photos turn awkward, and localized edits often spill into areas you never wanted touched. Qwen-Image-Edit-2511 is not built from scratch it is an enhanced version of Qwen-Image-Edit-2509, which already made solid progress in instruction driven editing and single/multiple subject consistency.
Natural language image editing with better consistency, stronger control, and real-world usability without extra hassle. Based on feedback and real usage patterns, the research focus was clear:
Instruction-Driven Editing That Actually Listens- You can make simple touch ups or complex semantic changes using natural language and the model follows instructions more faithfully than before.
Strong Localized Control-Object replacement, background swaps, material changes, or fine-grained edits happen only where you want them no accidental spillover.
Built-In LoRA Support-Popular community-created LoRAs are already integrated into the base model. That means you can unlock creative styles and effects without extra tuning or setup.
Industrial & Engineering-Friendly From batch product design to material replacement in industrial components, the model is clearly optimized for practical, production-level tasks.
Enhanced Geometric Reasoning- The model now understands geometry better and can generate these elements directly with auxiliary construction lines, annotations, or design guides.
Installation
1. Update ComfyUI from Manager by selecting Update All if you are familiar to it. New user have to install ComfyUI.
2. Download Qwen Image Edit 2511 FP8/BF16 (qwen_image_edit_2511_bf16.safetensors or qwen_image_edit_2511 FP8 mixed) and save it into ComfyUI/models/diffusion_models directory.
For low VRAM users, download gguf variants:-
(a) Qwen Image Edit 2511 gguf by unsloth or
(b) Qwen Image Edit 2511 gguf by quanstack
Save it into your ComfyUI/models/unet folder. Make sure you have already installed ComfyUI-GGUF custom node by Author-city96.
If not done, install it from the Manager by selecting Custom Nodes Manager option. If you do not know what's GGUF models, you can get detailed overview from our model quantization tutorial.
3. All the other models (text encoder, VAE) will be same as we used for Qwen Image Edit 2509 model. You do not need to download again. But if you want then download them :
(a) Download Text encoder (qwen_2.5_vl_7b_fp8_scaled.safetensors) and save it into your ComfyUI/models/text_encoders directory.
(b) Download Vae (qwen_image_vae.safetensors) and save it into ComfyUI/models/vae folder.
4. Download Lightning Lora model (Qwen-Image-Edit-2511-Lightning-4steps-V1.0-bf16.safetensors) if you want then save it into your ComfyUI/models/loras directory.
5. That's it, just restart and refresh ComfyUI to take effect.
Workflow
1. Download the workflow (Qwen_Image_Edit_2511.json) from our Hugging face repository.
2. Drag and drop into ComfyUI.
(a) Load multiple images into load Image node. You can use 1-3 images to merge and get the single image. You can enable/ disable the node by right clicking on it and select Bypass.
(b) Load Qwen Image Edit 2511 model into Unet loader/Load diffusion model loader node.
(c) Load the Vae and text encoders.
(d) Set the KSampler Settings for different model variants:
BF16 variant-
CFG-4
Steps- 40
FP8 variant-
CFG-4.0
Steps- 20
FP8 + lighting lora variant-
CFG-1.0
Steps- 4
(e) Put the prompt into prompt box.
3 . Hit the Run button to edit your image.
It's not just about prettier images. It's about consistency, control, and workflow simplicity. The built in LoRA support alone removes a big barrier for creators, while the improved multi-person handling and industrial focus make it useful beyond casual editing.
