HiMed (English): leaderboard
English control for the HiMed (Jiang et al., 2026) suite: the mean accuracy over four English multiple-choice sets, MMLU-Pro Biology, MMLU-Pro Health, the GPQA-med subset (organic chemistry, molecular biology and genetics) and the English NEET-UG questions of HiMed-West Exam, each zero-shot multiple choice scored with a relaxed answer extractor, averaged over three runs. Accuracy in percent; higher is better.
Metric: Accuracy (%; mean of the four English sets, zero-shot multiple choice, mean of three runs). Source: arxiv.org. Saturation forecast: Estimated already saturated. 11 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Qwen 2.5 14B Instruct | 61.1 | #1745 |
| 2 | GPT-4o | 59.9 | #741 |
| 3 | Qwen 2.5 7B Instruct | 54.6 | #2583 |
| 4 | Llama 3.1 8B Instruct | 50.8 | #4277 |
| 5 | aya-23-8B | 32 | #6778 |
No result here: #1 Claude Fable 5.2, #2 Gemini 4 Carbon, #5 Claude Opus 5.5.
Interactive version: theaggregate.ai/benchmark?slug=himed-english · How It Works · Data refreshed daily, snapshot 2026-11-12.