Qwen-Image-Bench - Human Expert Rating - Creative Generation: 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 creative generation (expressing authentic creative intent and style) pillar on a 1 to 10 scale; higher is better.
Metric: Creative generation (expressing authentic creative intent and style) 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
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | GPT Image 2 | 8.78 | |
| 2 | Nano Banana 2 | 8.48 | |
| 3 | Nano Banana Pro | 8.46 | |
| 4 | GPT Image 1.5 | 7.25 | |
| 5 | Qwen Image 2.0 Pro | 6.61 | |
| 6 | Seedream 5.0 | 6.43 | |
| 7 | Seedream 4.5 | 5.4 | |
| 8 | Qwen-Image-2512 | 5.22 | |
| 9 | GPT Image 1 | 5.1 | |
| 10 | Seedream 4.0 | 4.59 | |
| 11 | Imagen 4.0 Ultra | 4.41 | |
| 12 | FLUX 2 Max | 4.36 | |
| 13 | Imagen 4.0 | 4.29 | |
| 14 | FLUX 2 Pro | 3.97 | |
| 15 | HunyuanImage 3.0 | 3.4 |
Interactive version: theaggregate.ai/benchmark?slug=qwen-image-bench-human-expert-rating-creative-generation · How It Works · Data refreshed daily, snapshot 2026-11-21.