KVBench (Detailed Caption) - Physics (English): leaderboard
KVBench (2026) tests whether text-to-image models can draw correct senior high-school physics content, such as diagrams and scientific illustrations, from English prompts given as a detailed caption that spells out the required visual content. Qwen2.5-VL-32B checks each image against a list of textbook-grounded constraints; the score is the percentage satisfied, higher is better.
Metric: Checklist score (%) on the physics prompts in English, each given as a detailed caption that spells out the required visual content: KVBench knowledge-intensive text-to-image prompts from over 30 senior high-school textbooks (150 per subject and language, each with a textbook reference image); Qwen2.5-VL-32B-Instruct answers a checklist of four to six binary questions (key objects, attributes, spatial relations, reasoning outcomes) about each generated image, and the score is the mean share of satisfied items, in percent; higher is better. Source: arxiv.org. Saturation forecast: Rough model projection: around 2026. 16 models tracked. Updated 7 Oct 2026.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Nano Banana Pro (Gemini 3 Pro Image) | 76.2 | |
| 2 | Gemini 2.5 Flash Image | 73.54 | |
| 3 | GPT-Image (KVBench checkpoint unspecified) | 73.18 | |
| 4 | Seedream 4.0 | 70.82 | |
| 5 | FLUX.2-dev | 65.54 | |
| 6 | FLUX 2 Max | 64.74 | |
| 7 | SD3.5-Large (KVBench GGUF build, quantization unspecified) | 48.96 | |
| 8 | Qwen-Image | 48.67 | |
| 9 | FLUX.1-Krea-dev | 47.74 | |
| 10 | HiDream (KVBench checkpoint unspecified) | 45.28 | |
| 11 | Z-Image | 41.4 | |
| 12 | FLUX.1-dev | 37.78 | |
| 13 | BAGEL-7B-MoT | 29.38 | #4668 |
| 14 | OmniGen2 | 23.02 | #4625 |
| 15 | Show-o2 | 14.7 |
Interactive version: theaggregate.ai/benchmark?slug=kvbench-detailed-caption-physics-english · How It Works · Data refreshed daily, snapshot 2026-11-22.