SPUR - Quantitative Reasoning: leaderboard
Metric: Accuracy (%) on SPUR quantitative 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 verify inter-group differences, effect sizes and significance by numerical comparison and calculation; higher is better. Source: arxiv.org. Saturation forecast: Around 2031. 20 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Claude 3.7 Sonnet (Thinking) | 59.96 |
| 2 | Gemini 3 Pro (Preview) | 58.9 |
| 3 | GLM-4.5V | 58.48 |
| 4 | Gemini 2.5 Pro (Preview 06-05) | 57.94 |
| 5 | Llama 4 Maverick | 57.02 |
| 6 | Ministral-3-14B-Instruct-2512 | 56.81 |
| 7 | GPT-5.1 | 56.36 |
| 8 | O4 Mini (High) | 56.36 |
| 9 | Qwen 3 VL 30B A3B (Thinking) | 54.17 |
| 10 | Seed-1.6 | 53.89 |
| 11 | Ministral-3-8B-Instruct-2512 | 53.53 |
| 12 | Qwen 2.5 VL 72B Instruct | 52.51 |
| 13 | Qwen 3 VL 30B A3B Instruct | 51.41 |
| 14 | Gemma 3 27B (IT) | 51.38 |
| 15 | InternVL3-78B | 51.06 |
Interactive version: theaggregate.ai/benchmark?slug=spur-quantitative-reasoning · How It Works · Data refreshed daily, snapshot 2026-10-07.