MedDialBench - Reluctant: leaderboard
Metric: Diagnostic accuracy (%) over 85 OSCE cases (each cell counts whole cases out of 85) after a free-form consultation with a reluctant patient (disclosure, moderate), the final diagnosis judged correct or not by Qwen3-Max; higher is better. Source: arxiv.org. Saturation forecast: Around January 2027. 5 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Gemini 3.1 Pro (Preview) | 87.1 |
| 2 | Claude Opus 4.6 | 80 |
| 3 | GPT-5.4 (Non-reasoning) | 80 |
| 4 | DeepSeek V3.2 (Non-reasoning) | 68.2 |
| 5 | Qwen 3.5 Plus (Non-reasoning) | 64.7 |
Interactive version: theaggregate.ai/benchmark?slug=meddialbench-reluctant · How It Works · Data refreshed daily, snapshot 2026-10-07.