Instructions to use Qwen/Qwen-Image-Bench with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use Qwen/Qwen-Image-Bench with Transformers:

# Use a pipeline as a high-level helper
from transformers import pipeline

pipe = pipeline("image-text-to-text", model="Qwen/Qwen-Image-Bench")
messages = [
    {
        "role": "user",
        "content": [
            {"type": "image", "url": "https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/p-blog/candy.JPG"},
            {"type": "text", "text": "What animal is on the candy?"}
        ]
    },
]
pipe(text=messages)

# Load model directly
from transformers import AutoProcessor, AutoModelForImageTextToText

processor = AutoProcessor.from_pretrained("Qwen/Qwen-Image-Bench")
model = AutoModelForImageTextToText.from_pretrained("Qwen/Qwen-Image-Bench")
messages = [
    {
        "role": "user",
        "content": [
            {"type": "image", "url": "https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/p-blog/candy.JPG"},
            {"type": "text", "text": "What animal is on the candy?"}
        ]
    },
]
inputs = processor.apply_chat_template(
	messages,
	add_generation_prompt=True,
	tokenize=True,
	return_dict=True,
	return_tensors="pt",
).to(model.device)

outputs = model.generate(**inputs, max_new_tokens=40)
print(processor.decode(outputs[0][inputs["input_ids"].shape[-1]:]))

Notebooks
Google Colab
Kaggle
Local Apps

vLLM

How to use Qwen/Qwen-Image-Bench with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "Qwen/Qwen-Image-Bench"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "Qwen/Qwen-Image-Bench",
		"messages": [
			{
				"role": "user",
				"content": [
					{
						"type": "text",
						"text": "Describe this image in one sentence."
					},
					{
						"type": "image_url",
						"image_url": {
							"url": "https://cdn.britannica.com/61/93061-050-99147DCE/Statue-of-Liberty-Island-New-York-Bay.jpg"
						}
					}
				]
			}
		]
	}'

Use Docker

docker model run hf.co/Qwen/Qwen-Image-Bench

SGLang

How to use Qwen/Qwen-Image-Bench with SGLang:

Install from pip and serve model

# Install SGLang from pip:
pip install sglang
# Start the SGLang server:
python3 -m sglang.launch_server \
    --model-path "Qwen/Qwen-Image-Bench" \
    --host 0.0.0.0 \
    --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "Qwen/Qwen-Image-Bench",
		"messages": [
			{
				"role": "user",
				"content": [
					{
						"type": "text",
						"text": "Describe this image in one sentence."
					},
					{
						"type": "image_url",
						"image_url": {
							"url": "https://cdn.britannica.com/61/93061-050-99147DCE/Statue-of-Liberty-Island-New-York-Bay.jpg"
						}
					}
				]
			}
		]
	}'

Use Docker images

docker run --gpus all \
    --shm-size 32g \
    -p 30000:30000 \
    -v ~/.cache/huggingface:/root/.cache/huggingface \
    --env "HF_TOKEN=<secret>" \
    --ipc=host \
    lmsysorg/sglang:latest \
    python3 -m sglang.launch_server \
        --model-path "Qwen/Qwen-Image-Bench" \
        --host 0.0.0.0 \
        --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "Qwen/Qwen-Image-Bench",
		"messages": [
			{
				"role": "user",
				"content": [
					{
						"type": "text",
						"text": "Describe this image in one sentence."
					},
					{
						"type": "image_url",
						"image_url": {
							"url": "https://cdn.britannica.com/61/93061-050-99147DCE/Statue-of-Liberty-Island-New-York-Bay.jpg"
						}
					}
				]
			}
		]
	}'

Docker Model Runner
How to use Qwen/Qwen-Image-Bench with Docker Model Runner:
```
docker model run hf.co/Qwen/Qwen-Image-Bench
```

Q-Judger

A fine-tuned judge model for evaluating text-to-image (T2I) generation quality. Built on top of Qwen3.6-27B, it scores generated images across 5 hierarchical dimensions using structured checklists and outputs JSON-formatted evaluation results.

Links

Resource	Link
📑 Paper	http://arxiv.org/abs/2605.28091
📊 Benchmark Dataset (HuggingFace)	https://huggingface.co/datasets/Qwen/Qwen-Image-Bench
📊 Benchmark Dataset (ModelScope)	https://www.modelscope.cn/datasets/Qwen/Qwen-Image-Bench
💻 GitHub	https://github.com/QwenLM/Qwen-Image-Bench
🧑‍⚖️ Q-Judger Model	https://huggingface.co/Qwen/Qwen-Image-Bench
🧑‍⚖️ Q-Judger Model	https://modelscope.cn/models/Qwen/Qwen-Image-Bench

Model Description

Q-Judger is a vision-language model fine-tuned specifically for automated evaluation of text-to-image generated images. Given a text prompt and a generated image, the model evaluates the image on fine-grained quality criteria organized in a 3-level hierarchy and outputs structured JSON scores.

Base Model: Qwen3.6-27B
Task: Image quality evaluation / judging
Input: Text prompt + generated image
Output: Structured JSON with per-dimension scores (0 = Fail, 1 = Pass, 2 = Excel, N/A)
Thinking Mode: Enabled — the model uses chain-of-thought reasoning before producing the final JSON output

Evaluation Dimensions

The model evaluates images across 5 top-level dimensions, each with multiple sub-dimensions:

Quality

Realism: Physical Logic, Material Texture
Detail: Noise, Edge Clarity, Naturalness
Resolution: Resolution

Aesthetics

Composition: Composition
Color Harmony: Color Harmony
Lighting: Lighting & Atmosphere
Anatomical Portraiture: Anatomical Fidelity
Emotional Expression: Emotional Expression
Style Control: Style Control

Alignment

Attributes: Quantity, Facial Expression, Material Properties, Color, Shape, Size
Actions: Contact Interaction, Non-contact Interaction, Full-body Action
Layout: 2D Space, 3D Space
Relations: Composition Relationship, Difference/Similarity, Containment
Scene: Real-world Scene, Virtual Scene

Real-world Fidelity

Fairness: Social Bias, Cultural Fairness
Safety & Compliance: Safety & Compliance
World Knowledge: Animals, Objects, Information Visualization, Temporal Characteristics, Cultural Elements

Creative Generation

Imagination: Imagination
Feature Matching: Feature Matching
Logical Resolution: Logical Resolution
Text Rendering: Text Accuracy, Text Layout, Font, Cross-lingual Generation
Design Applications: Graphic Design, Product Design, Spatial Design, Fashion Styling, Game Design, Art Design
Visual Storytelling: Cinematic Style, Camera / Lens Style, Storyboard Creation, Shot Sizes, Composition, Angles, Comic Creation

Scoring Methodology

Raw Score Mapping

Raw Score	Meaning	Mapped Score
0	Fail	0
1	Pass	60
2	Excel	100
N/A	Not applicable	Excluded

Aggregation

Level-3 → Level-2: Average all non-N/A Level-3 scores within a Level-2 category
Level-2 → Level-1: Average all Level-2 scores within a Level-1 dimension
Level-1 → Total: Average all Level-1 dimension scores

Human Agreement

We validate the judge model against human expert rankings by computing Spearman rank correlation ($\rho$) between the model's rankings and human expert rankings across the five L1 pillars and overall. All correlations are statistically significant ($p < 10^{-4}$, $N = 18$ models).

Dimension	Spearman $\rho$
Quality	0.89
Aesthetics	0.89
Alignment	0.89
Real-world Fidelity	0.92
Creative Generation	0.92
Overall	0.92

Quick Start

Get the Inference Code

git clone https://github.com/QwenLM/Qwen-Image-Bench.git
cd Qwen-Image-Bench

Installation

1. Create and activate a virtual environment with uv:

uv venv myenv --python 3.11
source myenv/bin/activate

2. Install PyTorch (select the command matching your CUDA version):

See the official guide: https://pytorch.org/get-started/locally/

3. Install Python dependencies:

uv pip install -r requirements.txt

This installs all required dependencies including ms-swift.

Run Inference

python judge.py \
  --input your_data.jsonl \
  --model Qwen/Qwen-Image-Bench

Input Format

Prepare a CSV, JSON, or JSONL file with the following columns:

Column	Type	Description
`ID`	int	Prompt identifier (1-1000), must match benchmark metadata
`prompt`	str	The text prompt used to generate the image
`image_path`	str	Path to the generated image file

Output Format

The model outputs a JSON object per dimension, structured as:

{
  "Level-2 Dimension": {
    "Level-3 Dimension": {"score": 0|1|2|"N/A"}
  }
}

Example (Quality dimension):

{
  "Realism": {
    "Physical Logic": {"score": 1},
    "Material Texture": {"score": 2}
  },
  "Detail": {
    "Noise": {"score": 1},
    "Edge Clarity": {"score": 1},
    "Naturalness": {"score": 1}
  },
  "Resolution": {
    "Resolution": {"score": 2}
  }
}

CLI Options

Argument	Default	Description
`--input`	(required)	Input CSV/JSON/JSONL with ID, prompt, image_path
`--model`	(required)	HuggingFace model ID or local model path
`--hf-bench-repo`	-	HF dataset repo for bench metadata
`--local-metadata`	-	Local metadata file path (overrides default)
`--max-batch-size`	24	ms-swift max_batch_size
`--max-new-tokens`	4096	Max generation tokens

Inference Parameters

The judge model uses fixed inference parameters for reproducibility:

Parameter	Value
`seed`	42
`temperature`	0
`top_k`	1
`top_p`	1.0
`repetition_penalty`	1.05
`max_new_tokens`	4096
`enable_thinking`	True
`max_batch_size`	24

Citation

If you find this model useful, please cite our paper:

@misc{li2026qwenimagebenchgenerationcreationtexttoimage,
      title={Qwen-Image-Bench: From Generation to Creation in Text-to-Image Evaluation}, 
      author={Niantong Li and Guangzheng Hu and Weixu Qiao and Ying Ba and Qichen Hong and Shijun Shen and Jinlin Wang and Fan Zhou and Jianye Kang and Xin Shang and Ziyi He and Wei Wang and Dalin Li and Jiahao Li and Jie Zhang and Kaiyuan Gao and Kun Yan and Lihan Jiang and Ningyuan Tang and Shengming Yin and Tianhe Wu and Xiao Xu and Xiaoyue Chen and Yuxiang Chen and Yan Shu and Yanran Zhang and Yilei Chen and Yixian Xu and Zekai Zhang and Zhendong Wang and Zihao Liu and Zikai Zhou and Hongzhu Shi and Yi Wang and Bing Zhao and Hu Wei and Lin Qu and Chenfei Wu},
      year={2026},
      eprint={2605.28091},
      archivePrefix={arXiv},
      primaryClass={cs.CV},
      url={https://arxiv.org/abs/2605.28091}, 
}