SPUR - Qualitative Reasoning: leaderboard

SPUR (2026) tests multimodal models with multiple-choice questions on multi-panel biomedical experimental figures from PubMed Central. Accuracy in percent on qualitative reasoning questions, which ask the model to interpret biological significance from visual evidence and experimental design; higher is better.

Metric: Accuracy (%) on SPUR qualitative reasoning multiple-choice questions about multi-panel biomedical experimental figures from PubMed Central papers (questions drafted by GPT-4o, kept only when GPT-4o failed them in at least six of ten text-only attempts, then expert-reviewed): the questions ask the model to combine visual evidence, domain knowledge and experimental design to interpret biological significance; higher is better. Source: arxiv.org. Saturation forecast: Estimated already saturated. 19 models tracked. Updated 7 Oct 2026.

Top models

#ModelSettingScoreOverall rank
1Gemini 3 Pro (Preview)90.31#119
2Claude 3.7 SonnetThinking87.58#470
3Gemini 2.5 Pro (Preview 06-05)86.54#213
4GPT-5.186.52#240
5Llama 4 Maverick84.64#1110
6O4 MiniHigh84.33#368
7GLM-4.5V80.94#754
8Seed-1.680.31#469
9InternVL3-78B75.24#759
10Qwen 3 VL 30B A3B (Thinking)75.16#719
11Grok 4.1 Fast73.44#427
12Qwen 2.5 VL 72B Instruct73.1#841
13Mistral Small 3.172.82#1706
14Ministral-3-14B-Instruct-251272.5#1639
15Ministral-3-8B-Instruct-251270.85#1907

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

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