Qwen-Image-Bench - Human Expert Rating - Quality: leaderboard

Qwen-Image-Bench (2026) scores text-to-image models on 1,000 creator-designed prompts in five pillars: quality, aesthetics, alignment, real-world fidelity and creative generation. This board is the professional annotators rating of the visual quality pillar on a 1 to 10 scale; higher is better.

Metric: Visual quality pillar professional annotator rating on a 1 to 10 scale over 1,000 prompts per pillar, averaged; higher is better. Source: arxiv.org. Saturation forecast: Rough model projection: around 2026. 18 models tracked. Updated 7 Oct 2026.

Top models

#ModelScoreOverall rank
1GPT Image 29.48
2Nano Banana 28.87
3Nano Banana Pro8.79
4GPT Image 1.57.37
5Qwen Image 2.0 Pro7.03
6Seedream 5.06.55
7Seedream 4.55.56
8Qwen-Image-25125.48
9GPT Image 15.2
10Seedream 4.04.75
11Imagen 4.0 Ultra4.74
12Imagen 4.04.67
13FLUX 2 Pro4.57
14FLUX 2 Max4.48
15HunyuanImage 3.03.28

Interactive version: theaggregate.ai/benchmark?slug=qwen-image-bench-human-expert-rating-quality · How It Works · Data refreshed daily, snapshot 2026-11-21.