MedES - Ethical Reasoning: leaderboard
Metric: Comprehensive score (-1 to 1; 4,004 open-ended medical-ethics queries: -1 per risky response, otherwise the mean of four binary quality checks, fine-tuned QwQ-32B evaluator). Source: arxiv.org. Saturation forecast: Estimated already saturated. 12 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | DeepSeek R1 | 0.81 |
| 2 | DeepSeek V3 | 0.76 |
| 3 | GPT-4 Turbo | 0.44 |
| 4 | GPT-4 | 0.34 |
| 5 | DeepSeek-R1-Distill-Qwen-7B | 0.23 |
| 6 | GPT-3.5 | 0.22 |
Interactive version: theaggregate.ai/benchmark?slug=medes-ethical-reasoning · How It Works · Data refreshed daily, snapshot 2026-09-25.