EHRBench: leaderboard
Metric: Overall Acc (%) (self-reported). Source: benchmarklist.com. 31 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | GPT-5.2 | 70.91 |
| 2 | GPT-4.1 | 69.43 |
| 3 | GPT-5 | 69.06 |
| 4 | Llama 3.3 70B | 67.28 |
| 5 | GPT-4.1 Mini | 66.79 |
| 6 | Qwen 3 32B | 66.78 |
| 7 | GPT-5 Mini | 66.12 |
| 8 | GLM-4 32B | 66.12 |
| 9 | Qwen 2.5 32B | 64.97 |
| 10 | Llama 3 70B | 63.35 |
| 11 | Qwen 3 8B | 60.87 |
| 12 | Qwen 3 4B | 60.63 |
| 13 | GPT-4.1 Nano | 60.48 |
| 14 | glm-4-9B | 59.62 |
| 15 | Yi 1.5 34B | 58.94 |
Interactive version: theaggregate.ai/benchmark?slug=ehrbench · How It Works · Data refreshed daily, snapshot 2026-09-05.