ClinicalMC (English): leaderboard
Metric: Mean of 11 task scores (%; triage accuracy, examination recall, diagnosis F1, LLM-judged diagnosis basis and differential diagnosis, treatment and per-course IoU; GPT-4o-mini patient and examiner agents, gold inputs from earlier stages, mean of three runs; 5,804 English cases from PMC-Patients). Source: arxiv.org. Saturation forecast: Around 2032. 21 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Mixtral 8x22B Instruct (v0.1) | 49.8 |
| 2 | Qwen Turbo | 49.68 |
| 3 | GPT-5 Mini | 46.88 |
| 4 | GPT-4o Mini | 46.23 |
| 5 | Qwen 2.5 32B Instruct | 45.98 |
| 6 | Qwen 3 Next 80B A3B Instruct | 45.59 |
| 7 | Qwen 2.5 14B Instruct | 45.38 |
| 8 | Qwen 2.5 72B Instruct | 45.15 |
| 9 | Mistral 7B Instruct (v0.3) | 45.06 |
| 10 | Qwen 2.5 7B Instruct | 44.98 |
| 11 | Llama 3.3 70B Instruct | 44.52 |
| 12 | Llama 3.2 3B Instruct | 36.47 |
Interactive version: theaggregate.ai/benchmark?slug=clinicalmc-english · How It Works · Data refreshed daily, snapshot 2026-09-26.