Qwen-Image ComfyUI Native Workflow Example – Association of Applied Artificial Intelligence

Qwen-Image is a 20B parameter MMDiT (Multimodal Diffusion Transformer) model open-sourced under the Apache 2.0 license.

Qwen-Image is the first image generation foundation model released by Alibaba’s Qwen team. It’s a 20B parameter MMDiT (Multimodal Diffusion Transformer) model open-sourced under the Apache 2.0 license. The model has made significant advances in complex text rendering and precise image editing, achieving high-fidelity output for multiple languages including English and Chinese.Model Highlights:

Excellent Multilingual Text Rendering: Supports high-precision text generation in multiple languages including English, Chinese, Korean, Japanese, maintaining font details and layout consistency
Diverse Artistic Styles: From photorealistic scenes to impressionist paintings, from anime aesthetics to minimalist design, fluidly adapting to various creative prompts

Related Links:

Currently Qwen-Image has multiple ControlNet support options available:

Qwen-Image-DiffSynth-ControlNets/model_patches: Includes canny, depth, and inpaint models
qwen_image_union_diffsynth_lora.safetensors: Image structure control LoRA supporting canny, depth, pose, lineart, softedge, normal, openpose
InstantX ControlNet: To be updated

ComfyOrg Qwen-Image live stream

Qwen-Image in ComfyUI – Lightning & LoRAshttps://www.youtube.com/embed/WBFHwrpYRtY?si=uREGRhBDryTJBIryQwen-Image ControlNet in ComfyUI – DiffSynthhttps://www.youtube.com/embed/bXMClHfEFn4?si=dcaNdqOMSwvu3t8x

Qwen-Image Native Workflow Example

Make sure your ComfyUI is updated.

Workflows in this guide can be found in the Workflow Templates. If you can’t find them in the template, your ComfyUI may be outdated.(Desktop version’s update will delay sometime)If nodes are missing when loading a workflow, possible reasons:

You are not using the latest ComfyUI version(Nightly version)
You are using Stable or Desktop version (Latest changes may not be included)
Some nodes failed to import at startup

There are three different models used in the workflow attached to this document:

Qwen-Image original model fp8_e4m3fn
8-step accelerated version: Qwen-Image original model fp8_e4m3fn with lightx2v 8-step LoRA
Distilled version: Qwen-Image distilled model fp8_e4m3fn

VRAM Usage Reference GPU: RTX4090D 24GB

Model Used	VRAM Usage	First Generation	Second Generation
fp8_e4m3fn	86%	≈ 94s	≈ 71s
fp8_e4m3fn with lightx2v 8-step LoRA	86%	≈ 55s	≈ 34s
Distilled fp8_e4m3fn	86%	≈ 69s	≈ 36s

1. Workflow File

After updating ComfyUI, you can find the workflow file in the templates, or drag the workflow below into ComfyUI to load it.

Download Workflow for Qwen-Image Official ModelDistilled versionDownload Workflow for Distilled Model

2. Model Download

Available Models in ComfyUI

Qwen-Image_bf16 (40.9 GB)
Qwen-Image_fp8 (20.4 GB)
Distilled versions (non-official, requires only 15 steps)

All models are available at Huggingface and ModelscopeDiffusion model

qwen_image_fp8_e4m3fn.safetensors

Qwen_image_distill

The original author of the distilled version recommends using 15 steps with cfg 1.0.
According to tests, this distilled version also performs well at 10 steps with cfg 1.0. You can choose either euler or res_multistep based on the type of image you want.

LoRA

Qwen-Image-Lightning-8steps-V1.0.safetensors

Text encoder

qwen_2.5_vl_7b_fp8_scaled.safetensors

VAEqwen_image_vae.safetensorsModel Storage Location

Copy

Ask AI

📂 ComfyUI/
├── 📂 models/
│   ├── 📂 diffusion_models/
│   │   ├── qwen_image_fp8_e4m3fn.safetensors
│   │   └── qwen_image_distill_full_fp8_e4m3fn.safetensors ## 蒸馏版
│   ├── 📂 loras/
│   │   └── Qwen-Image-Lightning-8steps-V1.0.safetensors   ## 8步加速 LoRA 模型
│   ├── 📂 vae/
│   │   └── qwen_image_vae.safetensors
│   └── 📂 text_encoders/
│       └── qwen_2.5_vl_7b_fp8_scaled.safetensors

3. Workflow Instructions

Make sure the Load Diffusion Model node has loaded qwen_image_fp8_e4m3fn.safetensors
Make sure the Load CLIP node has loaded qwen_2.5_vl_7b_fp8_scaled.safetensors
Make sure the Load VAE node has loaded qwen_image_vae.safetensors
Make sure the EmptySD3LatentImage node is set with the correct image dimensions
Set your prompt in the CLIP Text Encoder node; currently, it supports at least English, Chinese, Korean, Japanese, Italian, etc.
If you want to enable the 8-step acceleration LoRA by lightx2v, select the node and use Ctrl + B to enable it, and modify the Ksampler settings as described in step 8
Click the Queue button, or use the shortcut Ctrl(cmd) + Enter to run the workflow
For different model versions and workflows, adjust the KSampler parameters accordingly

The distilled model and the 8-step acceleration LoRA by lightx2v do not seem to be compatible for simultaneous use. You can experiment with different combinations to verify if they can be used together.

Qwen Image InstantX ControlNet Workflow

This is a ControlNet model, so you can use it as normal ControlNet.

1. Workflow and Input Images

Download the image below and drag it into ComfyUI to load the workflow

Download JSON Format WorkflowDownload the image below as input

2. Model Links

InstantX Controlnet

Download Qwen-Image-InstantX-ControlNet-Union.safetensors and save it to the ComfyUI/models/controlnet/ folder

Lotus Depth model

We will use this model to generate the depth map of the image. The following two models need to be downloaded:Diffusion Model

lotus-depth-d-v1-1.safetensors

VAE Model

vae-ft-mse-840000-ema-pruned.safetensors or any SD1.5 VAE

Copy

Ask AI

ComfyUI/
├── models/
│   ├── diffusion_models/
│   │   └─── lotus-depth-d-v1-1.safetensors
│   └── vae/
│       └──  lvae-ft-mse-840000-ema-pruned.safetensors

You can also use custom nodes like comfyui_controlnet_aux to generate depth map.