MicroVQA - Expert Visual Understanding: leaderboard

Accuracy (%) on the MicroVQA questions of the expert visual understanding task: interpreting what a research microscopy image shows. Each model answers every question of this task once with the same chain-of-thought prompt as the headline board.

Metric: Accuracy (%). Source: jmhb0.github.io. Saturation forecast: Estimated already saturated. 14 models tracked. Updated 17 Nov 2026.

Top models

#ModelScoreOverall rank
1O1 (2024-12-17)55.4#314
2Qwen 2 VL 7B Instruct54.1#2594
3pixtral-large-241150.8#1284
4Grok 2 (1212)50.3#1082
5Qwen 2 VL 72B Instruct49.2#1382
6Gemini 1.5 Flash-8B (001)48.7#2311
7GPT-4o Mini (2024-07-18)48.5#1494
8Claude 3.5 Haiku (20241022)48#1589
9VILA1.5-40B47.2#3238
10Pixtral-12B-240946.9#2537
11Llama 3.2 90B Vision Instruct44.9#1791
12Llama 3.1 Nemotron 70B Instruct44.9#1471
13VILA1.5-13B41.8#7029
14Llama 3.2 11B Instruct32.4#5226

No result here: #1 Claude Fable 5.2, #2 GPT-6.2 Astra, #3 Gemini 4 Carbon.

Interactive version: theaggregate.ai/benchmark?slug=microvqa-expert-visual-understanding · How It Works · Data refreshed daily, snapshot 2026-11-17.