Appear2Meaning - Culture: leaderboard

Appear2Meaning (2026) tests whether vision-language models infer structured cultural metadata from an image of a museum object, over Getty and Metropolitan Museum objects from East Asia, the ancient Mediterranean, Europe and the Americas, judged against the museum records. Accuracy of the predicted culture field in percent; higher is better.

Metric: Culture attribute accuracy (%, times 100; judged correct) on the Appear2Meaning benchmark (750 Getty and Met museum objects, 50 per culture-type combination over four cultural regions and four object types; image-only input, structured JSON prediction; GPT-4.1-mini judge against the museum metadata); higher is better. Source: arxiv.org. Saturation forecast: Around 2029. 9 models tracked. Updated 7 Oct 2026.

Top models

#ModelScoreOverall rank
1Qwen3-VL-Flash36.7#734
2Qwen 3 VL Plus35.3#514
3Qwen-VL-Max33.6#1402
4GPT-4.1 Mini33.1#799
5GPT-5.4 Mini33.1#409
6Qwen 3 VL 32B Instruct29.6#607
7Qwen 3 VL 8B Instruct26.8#1043
8Claude Haiku 4.524.9#587
9Pixtral-12B23.7#2611

No result here: #2 Claude Fable 5.2, #3 Gemini 4 Neon, #4 GPT-6.2 Astra.

Interactive version: theaggregate.ai/benchmark?slug=appear2meaning-culture · How It Works · Data refreshed daily, snapshot 2026-11-22.