EHRBench - Treatment: leaderboard
EHRBench (2026) tests LLMs on clinical decision questions generated from real patient records and verified against biomedical knowledge bases, posed as multiple-choice questions with 4 to 6 options. This board is accuracy in percent on the treatment-decision questions; higher is better.
Metric: Accuracy (%) on the treatment-decision questions, EHR-grounded multiple-choice clinical decision questions (4, 5 and 6 options) built from MIMIC-III, MIMIC-IV and PROMOTE encounter trajectories with knowledge-base verification, JSON-constrained answers, deterministic decoding; higher is better. Source: arxiv.org. Saturation forecast: Estimated already saturated. 31 models tracked. Updated 7 Oct 2026.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | GPT-5 Chat | 80.45 | #449 |
| 2 | GPT-5.2 Instant | 80.13 | #304 |
| 3 | GPT-4.1 | 80.1 | #505 |
| 4 | Llama 3.3 70B Instruct | 79.05 | #1385 |
| 5 | GPT-4.1 Mini | 77.9 | #789 |
| 6 | GLM-4 32B (0414) | 77.9 | #1505 |
| 7 | EHRBench llama3-70B (checkpoint unspecified) | 77.63 | |
| 8 | Qwen 3 32B | 77.34 | #972 |
| 9 | Qwen 2.5 32B | 76.87 | #1344 |
| 10 | GPT-5 Mini | 76.17 | #364 |
| 11 | Mistral Small 3 | 75.02 | #1796 |
| 12 | m1-32B-1k | 74.84 | #1240 |
| 13 | Qwen 3 8B | 74.49 | #1873 |
| 14 | doctor-r1-8B | 74.49 | #1624 |
| 15 | GPT-4.1 Nano | 74.03 | #2056 |
No result here: #1 Claude Fable 5.2, #3 Gemini 4 Neon, #4 GPT-6.2 Astra.
Interactive version: theaggregate.ai/benchmark?slug=ehrbench-treatment · How It Works · Data refreshed daily, snapshot 2026-11-21.