Day-1 Support of Qwen-Image InstantX ControlNet

1 model with 4 controls: Canny, Soft Edge, Depth and Pose

The new Qwen-Image Unified ControlNet model just dropped! You can immediately take advantage of structure-guided generation inside ComfyUI with the latest advancements from the Qwen-Image model.

Model Highlights

This model provides a unified ControlNet designed specifically for Qwen-Image.
It supports four common control types in a single model:

• Canny – edge detection for precise outlines

• Soft Edge – smoother, more flexible structural guidance

• Depth – 3D-aware generation with depth maps

• Pose – human keypoint guidance for body and action control

Get Started

1. Update ComfyUI or ComfyUI desktop to the latest

2. Use our template workflow below.

3. Download the models as guided in the workflow

4. Start creating!

Examples

Qwen Image + InstantX ControlNet - Depth

Qwen Image + InstantX ControlNet - Canny

Qwen Image + InstantX ControlNet - OpenPose

Qwen Image + InstantX ControlNet with subgraphs
You can now also use subgraphs to clean up the workflow canvas and easily compare the results from different inputs.

Check our documentation for more details: https://docs.comfy.org/tutorials/image/qwen/qwen-image#qwen-image-instantx-controlnet-workflow

As always, enjoy creating!