KVBench (Detailed Caption) - Chemistry (Chinese): leaderboard

KVBench (2026) tests whether text-to-image models can draw correct senior high-school chemistry content, such as diagrams and scientific illustrations, from Chinese 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 chemistry prompts in Chinese, 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

#ModelScoreOverall rank
1Nano Banana Pro (Gemini 3 Pro Image)82.95
2GPT-Image (KVBench checkpoint unspecified)54.09
3Seedream 4.050.32
4Gemini 2.5 Flash Image39.01
5FLUX.2-dev33.06
6Qwen-Image31.01
7Z-Image30.04
8FLUX 2 Max24.76
9OmniGen212.31#4579
10HiDream (KVBench checkpoint unspecified)11.08
11BAGEL-7B-MoT10.59#4619
12Janus-Pro9.86
13FLUX.1-Krea-dev8.67
14SD3.5-Large (KVBench GGUF build, quantization unspecified)7.86
15FLUX.1-dev6.21

Interactive version: theaggregate.ai/benchmark?slug=kvbench-detailed-caption-chemistry-chinese · How It Works · Data refreshed daily, snapshot 2026-11-21.