Qwen-Image-Bench - Aesthetics: 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 aesthetics pillar score from the judge model on the Chinese prompts, 0 to 100; higher is better.

Metric: Aesthetics pillar Q-Judger score (0-100), the unified judge model fine-tuned from Qwen3.6-27B on expert annotations, aggregated bottom-up from 56 third-level facets, over the 1,000 creator-designed Chinese prompts; 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 267.53
2Nano Banana 261.08
3GPT Image 1.560.88
4Nano Banana Pro60.26
5Seedream 4.058.81
6Seedream 4.558.72
7Qwen Image 2.0 Pro58.67
8Seedream 5.058.4
9FLUX 2 Pro56.94
10FLUX 2 Max56.85
11GPT Image 155.09
12Qwen-Image-251254.74
13Imagen 4.0 Ultra54.25
14HunyuanImage 3.053.57
15Imagen 4.052.68

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