SPUR - Numerical Perception: leaderboard
SPUR (2026) tests multimodal models with multiple-choice questions on multi-panel biomedical experimental figures from PubMed Central. Accuracy in percent on numerical perception questions, which ask the model to quantify levels and differences of visual features within a panel; higher is better.
Metric: Accuracy (%) on SPUR the 636 numerical perception 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 quantify absolute levels and differences of visual features within a panel; higher is better. Source: arxiv.org. Saturation forecast: Around August 2027. 19 models tracked. Updated 7 Oct 2026.
Top models
| # | Model | Setting | Score | Overall rank |
|---|---|---|---|---|
| 1 | Gemini 3 Pro (Preview) | 61.26 | #119 | |
| 2 | Claude 3.7 Sonnet | Thinking | 59.67 | #470 |
| 3 | O4 Mini | High | 59.24 | #368 |
| 4 | GPT-5.1 | 58.73 | #240 | |
| 5 | GLM-4.5V | 57.7 | #754 | |
| 6 | Gemini 2.5 Pro (Preview 06-05) | 56.47 | #213 | |
| 7 | Qwen 3 VL 30B A3B (Thinking) | 53.48 | #719 | |
| 8 | Seed-1.6 | 53.3 | #469 | |
| 9 | Llama 4 Maverick | 51.73 | #1110 | |
| 10 | Ministral-3-8B-Instruct-2512 | 51.57 | #1907 | |
| 11 | Ministral-3-14B-Instruct-2512 | 50.88 | #1639 | |
| 12 | Grok 4.1 Fast | 47.33 | #427 | |
| 13 | InternVL3-78B | 46.3 | #759 | |
| 14 | Qwen 3 VL 30B A3B Instruct | 44.18 | #895 | |
| 15 | Mistral Small 3.1 | 42.74 | #1706 |
No result here: #1 Claude Fable 5.2, #3 Gemini 4 Neon, #4 GPT-6.2 Astra.
Interactive version: theaggregate.ai/benchmark?slug=spur-numerical-perception · How It Works · Data refreshed daily, snapshot 2026-11-21.