MM-NeuroOnco (Closed-Ended) - Shape: leaderboard
Metric: Accuracy (%) on lesion shape regularity; five-option multiple choice ('None of the above' appended to every question as a rejection option, chance 20%) on 2D brain-tumor MRI slices from the 1,000-image MM-NeuroOnco-Bench, ground truth from the source datasets' labels and segmentation masks only; higher is better. Source: arxiv.org. Saturation forecast: Around 2036. 10 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Gemini 3 Flash | 30.8 | #93 |
| 2 | Claude Sonnet 4 | 28.3 | #194 |
| 3 | MedGemma 1.5 4B | 28 | #775 |
| 4 | GPT-5.1 | 21.2 | #131 |
No result here: #3 Claude Opus 5.5, #5 GPT-6 Astra, #8 Claude Fable 5.1.
Interactive version: theaggregate.ai/benchmark?slug=mm-neuroonco-closed-ended-shape · How It Works · Data refreshed daily, snapshot 2026-10-11.