MedS-Bench - Named Entity Recognition: leaderboard
The Named Entity Recognition subtask of MedS-Bench (arXiv) — tagging clinical entities in text; F1 (%). Published overview values are retained where detailed tables differ.
Metric: F1 (%). Source: henrychur.github.io. 11 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | GPT-4 | 59.52 | #438 |
| 2 | Claude 3.5 Sonnet (20240620) | 25.56 | #299 |
Interactive version: theaggregate.ai/benchmark?slug=meds-bench-named-entity-recognition · How It Works · Data refreshed daily, snapshot 2026-10-20.