HealthGPT-Pro-4B: benchmark results
Provider: Other.
Aggregate Rating 1508 ± 52, rank #2175 of 10552 rated models, from 12 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| ObGynLongBench - History-Level EHR - L3 Long-Horizon Reasoning | 57.3 | Accuracy (%) | 59.4 |
| MedPIC-Bench - Rule Deactivation | 41.2 | Accuracy (%) | 55.6 |
| ObGynLongBench - History-Level EHR | 57.1 | Accuracy (%; full pre-decision EHR history) | 25 |
| ObGynLongBench - History-Level EHR - L1 Single Evidence | 59 | Accuracy (%) | 25 |
| ObGynLongBench - History-Level EHR - L2 Multi-Source Integration | 56.1 | Accuracy (%) | 25 |
| ObGynLongBench - Visit-Level EHR | 54.1 | Accuracy (%; same-day visit records) | 25 |
| MedPIC-Bench - Counterfactual Pair | 11.2 | Both-correct pair rate (%) | 24.1 |
| MedPIC-Bench - Counterfactual | 33.3 | Accuracy (%) | 22.2 |
| ObGynLongBench - Evidence-Only | 69.6 | Accuracy (%; supporting evidence only) | 18.8 |
| MedPIC-Bench - Rule Activation | 36.6 | Accuracy (%) | 1.9 |
| MedPIC-Bench - Guideline-Following | 37.7 | Accuracy (%) | 0 |
| MedPIC-Bench - Overall | 36 | Accuracy (%) | 0 |
Interactive version: theaggregate.ai/model?slug=healthgpt-pro-4b · How It Works · Data refreshed daily, snapshot 2026-11-16.