EHR-Complex: leaderboard
Clinical agents promise to democratize access to electronic health records (EHRs), yet existing benchmarks fail to reflect the complexity of practical EHR analysis, e.g., often operating on idealized, clean EHRs via static SQL generation rather than interactive execution. In this work, we introduce EHR-Complex, a large-scale benchmark designed for interactive clinical database reasoning. Built on
Metric: Avg. (self-reported). Source: benchmarklist.com. Saturation forecast: Around April 2027. 14 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Kimi K2.5 | 62.3 | #144 |
| 2 | Qwen 3.5 397B A17B | 62.2 | #146 |
| 3 | DeepSeek V3.2 Exp | 59.2 | #234 |
| 4 | DeepSeek V3.1 | 55.9 | #264 |
| 5 | Qwen 3 235B A22B | 53.4 | #313 |
| 6 | GPT-4.1 Mini | 49.2 | #355 |
| 7 | GPT-4.1 | 46.8 | #248 |
| 8 | Qwen 3 32B | 35.7 | #443 |
| 9 | GPT-4o | 30.7 | #343 |
| 10 | Gemini 2.5 Pro | 30.7 | #152 |
| 11 | Qwen 3 14B | 30 | #537 |
| 12 | Qwen 3 4B | 16 | #829 |
No result here: #3 Claude Opus 5.5, #6 GPT-6 Astra, #7 GPT-6.1 Sol.
Interactive version: theaggregate.ai/benchmark?slug=ehr-complex · How It Works · Data refreshed daily, snapshot 2026-10-20.