Installing Qwen Image Edit 2509 offline with ComfyUI

Clothing & Accessories Edit Example

How to Set Up the Qwen Image Edit 2509 Model

1. Open your browser and access the Hugging Face repo for Qwen Image Edit.
2. Choose the model version based on your PC specs: FP8 (20GB) for 24GB VRAM, or GGUF (7GB) for 8GB VRAM.
3. Click the 'Download' button on the selected model file.
4. Move the downloaded model file into the folder: ComfyUI > models > diffusion_models.
5. Open ComfyUI and load your Qwen Image Edit workflow.
6. Locate the node named 'Load Diffusion Model' at the top of the workflow.
7. Click the dropdown menu within that node and select the '2509' model version you just downloaded.
8. In the 'Load CLIP' node, select the corresponding CLIP model (usually included in the repo).
9. In the 'Load VAE' node, ensure you select the correct VAE model for image decoding.

Performing Image Editing & Speed Optimization

1. Go to the 'Load Image' node and click the 'Upload' button to select the original image you want to edit.
2. Write your transformation instructions in the Text Prompt node (e.g., change clothing or accessories).
3. For maximum speed, find the 'Qwen Image Lightning' node and ensure it is enabled (Turn On).
4. If Lightning is enabled, go to the 'K-Sampler' node and change the 'Step Count' from 20 to 4 steps.
5. Click the 'Queue Prompt' button to start the image generation process.
6. View the final result in the 'Preview Image' or 'Save Image' node.