DiningBench - Carbohydrate Estimation: leaderboard

DiningBench (2026) evaluates vision-language models on multi-view user photos of restaurant dishes, testing dish classification against same-menu distractors, nutrition estimation and food visual question answering. This board is the mean absolute percentage error of the estimated carbohydrates; lower is better.

Metric: Mean absolute percentage error (%) of the estimated carbohydrates over the 1,650 DiningBench nutrition-estimation dishes (multi-view restaurant photos; ground truth from merchant labels or ingredient-based estimates cross-checked against USDA FoodData Central); lower is better. Source: arxiv.org. Saturation forecast: Around March 2027. 29 models tracked. Updated 7 Oct 2026.

Top models

#ModelScoreOverall rank
1Gemini 3 Pro (Preview)27.4#119
2Gemini 3 Flash (Preview)27.94#152
3GPT-4.135.81#505
4GPT-536.63#178
5O4 Mini37.38#368
6GPT-4o40.19#751
7Qwen 3 VL 30B A3B Instruct41.53#895
8Qwen 2.5 VL 72B Instruct41.7#841
9Gemini 2.5 Flash42.62#463
10Gemini 2.5 Pro43.01#276
11Qwen 2.5 VL 32B Instruct43.92#1062
12Qwen 3 VL 8B Instruct43.99#1029
13GPT-4o Mini45.03#1667
14Qwen 3 VL 30B A3B (Thinking)48.12#719
15MiMo-VL-7B-RL48.24#1155

No result here: #1 Claude Fable 5.2, #3 Gemini 4 Neon, #4 GPT-6.2 Astra.

Interactive version: theaggregate.ai/benchmark?slug=diningbench-carbohydrate-estimation · How It Works · Data refreshed daily, snapshot 2026-11-21.