🧠 Smart Setup for Qwen-Image-Edit-2509 in ComfyUI

UPDATE 10th November 2025 - Model Sampling Aura Flow Error.

It seems the recent ComfyUI Update also broke the nodes from this workflow here, not only from Wan 2.2. I updated all 4 workflows (2 for free members and 2 slightly improved versions for paid users you can find in the Bonus Link). Everything should work fine now!

BR -Chris

P.s.: If any problems arise, feel free to DM me anytime! I will get back to you asap. You are the backbone of my channel. Thanks for your support!

Original Post:

✨ New video’s out! I just posted a beginner-friendly setup guide for Qwen-Image-Edit-2509 inside ComfyUI — perfect for anyone looking to unlock open-source image editing that feels like Photoshop, but runs locally, for free, and on your GPU.

💝 BONUS: My Patreon supporters get an exclusive 1-click installer that sets this all up automatically. No terminal. No pain. Just results.
https://www.patreon.com/posts/qwen-image-edit-141073892

I’ve collected everything you need to get started — the exact model files, where to put them, and which ones to pick depending on your GPU's VRAM.

🎬 Watch the full video here:
📺 https://www.youtube.com/@AIKnowledge2Go

💻 Software Prerequisites

1. Install Git – https://git-scm.com/

2. Download ComfyUI – https://github.com/comfyanonymous/ComfyUI (50XX Series users: unzip directly, no install)

3. ComfyUI Manager – Open CMD in ComfyUI\custom_nodes and run:
   git clone https://github.com/Comfy-Org/ComfyUI-Manager.git

🧩 Must-Have Custom Nodes

KJNodes – https://github.com/kijai/ComfyUI-KJNodes

rgthree-comfy – https://github.com/rgthree/rgthree-comfy

ComfyUI-GGUF – https://github.com/city96/ComfyUI-GGUF

cg-use-everywhere – https://github.com/chrisgoringe/cg-use-everywhere

🧠 GGUF Base Model

This is the core image editing brain — pick based on your VRAM!
📂 Place it in: models\diffusion_models

Download from: https://huggingface.co/QuantStack/Qwen-Image-Edit-2509-GGUF/tree/main

VRAM Guide:
• 8 GB → qwen-image-edit-2509.Q4_K_M.gguf → fallback Q3_K_M
• 12 GB → Q5_K_M → fallback Q4_K_M
• 16 GB → Q6_K → fallback Q5_K_M
• 24 GB → Q6_K
• 32 GB+ → Q8_0

⚡ Lightning LoRA

Optional, but recommended — these give Qwen image generation superpowers.
📂 Place in: models\loras

8 Steps (Better quality):
https://huggingface.co/lightx2v/Qwen-Image-Lightning/resolve/main/Qwen-Image-Lightning-8steps-V2.0.safetensors?download=true

4 Steps (Faster):
https://huggingface.co/lightx2v/Qwen-Image-Lightning/resolve/main/Qwen-Image-Lightning-4steps-V2.0.safetensors?download=true

Full repo: https://huggingface.co/lightx2v/Qwen-Image-Lightning/tree/main

🧬 Text Encoder

This model helps the AI understand your prompts more precisely.
📂 Place in: models\text_encoders

Download: https://huggingface.co/Comfy-Org/Qwen-Image_ComfyUI/resolve/main/split_files/text_encoders/qwen_2.5_vl_7b_fp8_scaled.safetensors?download=true

🖼️ VAE – Visual AutoEncoder

Handles how images are decoded and displayed.
📂 Place in: models\vae

Download: https://huggingface.co/Comfy-Org/Qwen-Image_ComfyUI/resolve/main/split_files/vae/qwen_image_vae.safetensors?download=true

🪄 That’s it!

You now have everything you need to get Qwen-Image-Edit-2509 working in ComfyUI like a pro.

✨ Join AIKnowledgeCentral on Patreon and get the full 1-click installer:
https://www.patreon.com/AIKnowledgeCentral

Let me know what you make with it! 💬
— Chris