GeoMMBench - Thermal Imagery: leaderboard

The thermal infrared imagery 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 thermal imagery 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 Flash75#723
2Gemini 1.5 Pro75#1063
3GPT-4o67.9#751
4O167.9#403
5Qwen 3 VL 8B67.9#917
6Qwen 2.5 VL 3B60.7#2639
7Qwen 3 VL 4B60.7#1151
8Qwen 3 VL 30B60.7#561
9Qwen 2.5 VL 7B57.1#1862
10InternVL2.5-8B53.6#2370
11LLaVA-NeXT-7B53.6#5032
12CogVLM253.6
13Qwen 2 VL 7B50#2275
14Phi-4 Multimodal Instruct50#4154
15InternLM-XComposer2.550#3817

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

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