qwen image ControlNet Redraw the image — v1.0
This is a workflow for qwen image using the pose, depth, hard edge, and soft edge features of ControlNet
Runtime profile
- Access
- External workflow
- Status
- Listed
- Version
- 千问洗图神器Qwen-Image-ControlNet-Union.json
- Source
- Civitai
- Updated
- Oct 17, 2026
Source description
This is a workflow for qwen image using the pose, depth, hard edge, and soft edge features of ControlNet
Usage 1、Upload the image for which you want to reverse-engineer the prompt 2、Upload the image for controlling pose, depth, and edges 3、Select control mode 4、Fill in the control strength [default 0.8, if the output effect is not ideal, you can appropriately increase the strength] 5、Choose whether to use the prompt words you enter, turn it on not to use the prompt words for reverse inference of images 6、Fill in your own prompt words [if using reverse image prompt words, you don't need to fill this]
You can click on 【Test Effect】 and it only takes about 4 minutes throughout. If you find the effect satisfactory, you can then download the workflow to deploy locally!
What you need
- Model files
- 37.4 GiB known subtotal
Remaining hardware or download-size estimates are not available yet.
Requirements
Models
9depth_anything_vitl14.pth
PossibleCheckpoint · Unknown
Checkpoint · 129 MB · PT · Unknown
InstantX_Qwen-Image-ControlNet-Union.safetensors
PossibleControlNet · Unknown
Text encoder · 8.74 GB · SAFETENSORS · fp8 · pruned · Hugging Face · Comfy-Org/Qwen-Image_ComfyUI
- File unverified
VAE · Hugging Face · Comfy-Org/Qwen-Image_ComfyUI
Qwen-Image-Lightning-8steps-V1.0.safetensors
File unverifiedLoRA · Hugging Face · lightx2v/Qwen-Image-Lightning
- Found
Checkpoint · 20.3 GB · GGUF · Unknown
Checkpoint · 8.23 GB · Hugging Face · nguoidoncui/ComfyUI
yolox_l.torchscript.pt
Not resolvedObject Detector · Unknown
Sources
1 source
Sources
1 sourceSource excerpts
1 excerpt
Source excerpts
1 excerptSource context: 182 downloads · Type Workflows · Base model Qwen
