CulturalMenuBench - Sub-Region Classification: leaderboard
CulturalMenuBench (arXiv 2609.03526v1): culinary multimodal reasoning over 587 dishes from 18 regions with dish and step-by-step cooking images, ingredients and procedure text, prompted in each cuisine's language. Table 7: four-way classification of Chinese dishes by regional cuisine from dish and cooking-step images for eight models; accuracy percent, chance 25, higher is better.
Metric: Accuracy (%; four-way Chinese regional cuisine, chance 25). Source: arxiv.org. Saturation forecast: Around December 2027. 8 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Claude Opus 4.6 | 56.1 | #67 |
| 2 | Kimi K2.6 | 56.1 | #113 |
| 3 | GPT-5.1 (2025-11-13) | 48.9 | #107 |
| 4 | Qwen 3.6 Plus | 44.2 | #124 |
| 5 | Qwen 3.5 35B A3B | 43.4 | #268 |
| 6 | GPT-5 Mini (2025-08-07) | 43.4 | #175 |
| 7 | GPT-5 Nano | 42.9 | #439 |
| 8 | Gemma 4 26B A4B (IT) | 38.1 | #293 |
No result here: #3 Claude Opus 5.5, #6 GPT-6 Astra, #7 GPT-6.1 Sol.
Interactive version: theaggregate.ai/benchmark?slug=culturalmenubench-sub-region-classification · How It Works · Data refreshed daily, snapshot 2026-11-05.