> ## Documentation Index
> Fetch the complete documentation index at: https://utopia.mintlify.site/llms.txt
> Use this file to discover all available pages before exploring further.

# Qwen-Image ComfyUI Native Workflow Example

> 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**:

* [GitHub](https://github.com/QwenLM/Qwen-Image)
* [Hugging Face](https://huggingface.co/Qwen/Qwen-Image)
* [ModelScope](https://modelscope.cn/models/qwen/Qwen-Image)

## Qwen-Image Native Workflow Example

<Tip>
  Make sure your ComfyUI is updated.

  * [Download ComfyUI](https://www.comfy.org/download)
  * [Update Guide](/installation/update_comfyui)

  Workflows in this guide can be found in the [Workflow Templates](/interface/features/template).
  If you can't find them in the template, your ComfyUI may be outdated.(Desktop version's update will delay sometime)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:

  1. Not using the latest ComfyUI version(Nightly version)
  2. Using Stable or Desktop version (Latest changes may not be included)
  3. Some nodes failed to import at startup
</Tip>

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

1. Qwen-Image original model fp8\_e4m3fn
2. 8-step accelerated version: Qwen-Image original model fp8\_e4m3fn with lightx2v 8-step LoRA
3. 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.
![Qwen-image Text-to-Image Workflow](https://raw.githubusercontent.com/Comfy-Org/example_workflows/main/image/qwen/qwen-image.png)

<a className="prose" target="_blank" href="https://raw.githubusercontent.com/Comfy-Org/workflow_templates/refs/heads/main/templates/image_qwen_image.json" style={{ display: 'inline-block', backgroundColor: '#0078D6', color: '#ffffff', padding: '10px 20px', borderRadius: '8px', borderColor: "transparent", textDecoration: 'none', fontWeight: 'bold'}}>
  <p className="prose" style={{ margin: 0, fontSize: "0.8rem" }}>Download Workflow for Qwen-Image Official Model</p>
</a>

Distilled version

<a className="prose" target="_blank" href="https://raw.githubusercontent.com/Comfy-Org/example_workflows/main/image/qwen/image_qwen_image_distill.json" style={{ display: 'inline-block', backgroundColor: '#0078D6', color: '#ffffff', padding: '10px 20px', borderRadius: '8px', borderColor: "transparent", textDecoration: 'none', fontWeight: 'bold'}}>
  <p className="prose" style={{ margin: 0, fontSize: "0.8rem" }}>Download Workflow for Distilled Model</p>
</a>

### 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](https://huggingface.co/Comfy-Org/Qwen-Image_ComfyUI/tree/main) and [Modelscope](https://modelscope.cn/models/Comfy-Org/Qwen-Image_ComfyUI/files)

**Diffusion model**

* [qwen\_image\_fp8\_e4m3fn.safetensors](https://huggingface.co/Comfy-Org/Qwen-Image_ComfyUI/resolve/main/split_files/diffusion_models/qwen_image_fp8_e4m3fn.safetensors)

Qwen\_image\_distill

* [qwen\_image\_distill\_full\_fp8\_e4m3fn.safetensors](https://huggingface.co/Comfy-Org/Qwen-Image_ComfyUI/resolve/main/non_official/diffusion_models/qwen_image_distill_full_fp8_e4m3fn.safetensors)
* [qwen\_image\_distill\_full\_bf16.safetensors](https://huggingface.co/Comfy-Org/Qwen-Image_ComfyUI/resolve/main/non_official/diffusion_models/qwen_image_distill_full_bf16.safetensors)

<Note>
  - 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.
</Note>

**LoRA**

* [Qwen-Image-Lightning-8steps-V1.0.safetensors](https://huggingface.co/lightx2v/Qwen-Image-Lightning/resolve/main/Qwen-Image-Lightning-8steps-V1.0.safetensors)

**Text encoder**

* [qwen\_2.5\_vl\_7b\_fp8\_scaled.safetensors](https://huggingface.co/Comfy-Org/Qwen-Image_ComfyUI/resolve/main/split_files/text_encoders/qwen_2.5_vl_7b_fp8_scaled.safetensors)

**VAE**

[qwen\_image\_vae.safetensors](https://huggingface.co/Comfy-Org/Qwen-Image_ComfyUI/resolve/main/split_files/vae/qwen_image_vae.safetensors)

**Model Storage Location**

```
📂 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. Complete the Workflow Step by Step

<img src="https://mintcdn.com/utopia/4h1gYSxdtAFmNpNH/images/tutorial/image/qwen/image_qwen_image-guide.jpg?fit=max&auto=format&n=4h1gYSxdtAFmNpNH&q=85&s=a9d366a9a0135389fff157e66dbbb991" alt="Step Guide" width="3111" height="1829" data-path="images/tutorial/image/qwen/image_qwen_image-guide.jpg" />

1. Make sure the `Load Diffusion Model` node has loaded `qwen_image_fp8_e4m3fn.safetensors`
2. Make sure the `Load CLIP` node has loaded `qwen_2.5_vl_7b_fp8_scaled.safetensors`
3. Make sure the `Load VAE` node has loaded `qwen_image_vae.safetensors`
4. Make sure the `EmptySD3LatentImage` node is set with the correct image dimensions
5. Set your prompt in the `CLIP Text Encoder` node; currently, it supports at least English, Chinese, Korean, Japanese, Italian, etc.
6. 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
7. Click the `Queue` button, or use the shortcut `Ctrl(cmd) + Enter` to run the workflow
8. For different model versions and workflows, adjust the KSampler parameters accordingly

<Note>
  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.
</Note>