GeoMMBench - Data Quality: leaderboard

The data quality questions (image quality inspection and radiometric, atmospheric and geometric correction) of GeoMMBench (2026), an expert-written multimodal geoscience and remote sensing question benchmark answered zero-shot (test set). Micro-averaged accuracy in percent; higher is better.

Metric: Micro-averaged accuracy (%) on the data quality questions (image quality inspection and atmospheric, radiometric and geometric corrections) of the GeoMMBench test set (expert-written multimodal geoscience and remote sensing multiple-choice and numeric questions, zero-shot, LMMs-Eval, rule-based answer extraction, invalid outputs counted wrong); higher is better. Source: arxiv.org. Saturation forecast: Around 2029. 35 models tracked. Updated 7 Oct 2026.

Top models

#ModelScoreOverall rank
1GPT-4o69.2#751
2Gemini 2.0 Flash69.2#723
3Phi-4 Multimodal Instruct65.4#4154
4LLaVA-OneVision-7B65.4#2981
5Qwen 3 VL 30B65.4#561
6Qwen 2 VL 2B61.5#6157
7Gemini 1.5 Pro61.5#1063
8Qwen 3 VL 8B57.7#917
9InternVL2.5-8B57.7#2370
10CogVLM257.7
11Qwen 3 VL 4B53.8#1151
12GPT-4V53.8#1470
13Qwen 2.5 VL 7B50#1862
14O150#403
15Qwen 2 VL 7B46.2#2275

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

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