Qwen-Image-Bench (English): 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 overall judge-model score on the English prompts, from 0 to 100, aggregated from 56 facets; higher is better.
Metric: Overall 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 English prompts (500 short, 500 long); higher is better. Source: arxiv.org. Saturation forecast: Rough model projection: around 2026. 18 models tracked. Updated 7 Oct 2026.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | GPT Image 2 | 65.23 | |
| 2 | GPT Image 1.5 | 60.42 | |
| 3 | Nano Banana 2 | 59.59 | |
| 4 | Nano Banana Pro | 59.33 | |
| 5 | Qwen Image 2.0 Pro | 57.9 | |
| 6 | Seedream 5.0 | 57.8 | |
| 7 | Seedream 4.5 | 56.95 | |
| 8 | Seedream 4.0 | 56.48 | |
| 9 | FLUX 2 Max | 56.14 | |
| 10 | FLUX 2 Pro | 55.61 | |
| 11 | GPT Image 1 | 54.24 | |
| 12 | Imagen 4.0 Ultra | 52.42 | |
| 13 | HunyuanImage 3.0 | 51.35 | |
| 14 | Qwen-Image-2512 | 51.32 | |
| 15 | Imagen 4.0 | 51.08 |
Interactive version: theaggregate.ai/benchmark?slug=qwen-image-bench-english · How It Works · Data refreshed daily, snapshot 2026-11-22.