GeoMMBench - GIS: leaderboard

The geographic information system discipline questions 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 geographic information system discipline questions 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 February 2027. 35 models tracked. Updated 7 Oct 2026.

Top models

#ModelScoreOverall rank
1Gemini 2.0 Flash86.4#723
2Gemini 1.5 Pro86.4#1063
3GPT-4o77.3#751
4Qwen 3 VL 8B77.3#917
5Qwen 3 VL 4B75#1151
6Qwen 2.5 VL 7B72.7#1862
7Qwen 3 VL 30B72.7#561
8Qwen 2 VL 7B70.5#2275
9O165.9#403
10LLaVA-OneVision-7B65.9#2981
11InternVL2.5-8B65.9#2370
12InternLM-XComposer2.563.6#3817
13Phi-3.5-Vision Instruct63.6#6268
14Phi-3-Vision-128K-Instruct63.6#7186
15Phi-4 Multimodal Instruct61.4#4154

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

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