VIPER (Veterinary Pathology) - KPrim: leaderboard
KPrim questions of VIPER (Harvard Medical School and Mass General Brigham, 2026): four true-false statements per item about a rat histology image, 100 when all four are correct, 50 for three, otherwise 0; mean in percent, higher is better.
Metric: KPrim score (%; four true-false statements per item, 100 for 4 of 4 correct, 50 for 3 of 4, else 0; the KPrim items of 1,251 questions written and validated by board-certified veterinary pathologists on 419 H&E-stained rat histology images from seven organ systems). Source: arxiv.org. Saturation forecast: Around February 2027. 16 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | GPT-5.4 | 54.3 | #123 |
| 2 | Gemma 4 31B (IT) | 54.1 | #342 |
| 3 | Claude Sonnet 4.6 | 47.1 | #161 |
| 4 | Qwen 3.5 27B | 46.6 | #359 |
| 5 | GPT-5.4 Mini | 45.9 | #389 |
| 6 | Gemini 2.5 Flash | 45 | #453 |
| 7 | GPT-5.4 Nano | 30.1 | #635 |
| 8 | MedGemma 1.5 4B | 18.4 | #2562 |
No result here: #1 Claude Fable 5.2, #2 Gemini 4 Carbon, #5 Claude Opus 5.5.
Interactive version: theaggregate.ai/benchmark?slug=viper-veterinary-pathology-kprim · How It Works · Data refreshed daily, snapshot 2026-11-13.