MediKAI: benchmark results
Provider: HelpingAI. Released 2024-04-01. Access: Open.
Aggregate Rating 1279 ± 38, rank #8992 of 10916 rated models, from 17 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| Open Medical LLM - PubMedQA | 74.2 | Accuracy (%) | 56.5 |
| Open LLM Leaderboard v1 - TruthfulQA MC2 | 48.77 | MC2 (%) (0-shot) | 42.6 |
| Open Medical LLM - MedMCQA | 45.71 | Accuracy (%) | 29.2 |
| Open LLM Leaderboard v1 - MMLU | 49.3 | Accuracy (%) (5-shot) | 28.5 |
| Open Medical LLM - MedQA (USMLE) | 47.29 | Accuracy (%) | 27.3 |
| Open Medical LLM - MMLU Anatomy | 50.37 | Accuracy (%) | 22.4 |
| Open LLM Leaderboard v1 - ARC Challenge | 46.5 | Normalized accuracy (%) (25-shot) | 21.6 |
| Open Medical LLM | 54.51 | Average Accuracy (%) | 21.4 |
| Open Medical LLM - MMLU Clinical Knowledge | 57.74 | Accuracy (%) | 21.4 |
| Open Medical-LLM Leaderboard (BenchmarkList) | 54.51 | Medical average (%) | 21.1 |
| Open Medical LLM - MMLU Professional Medicine | 48.53 | Accuracy (%) | 20.8 |
| Open Medical LLM - MMLU Medical Genetics | 60 | Accuracy (%) | 20.6 |
Interactive version: theaggregate.ai/model?slug=medikai · How It Works · Data refreshed daily, snapshot 2026-11-08.