Medmarks - AgentClinic — leaderboard
Metric: Score (%). Source: medmarks.ai. 12 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Gemini 3 Pro (Preview) | 85.9 |
| 2 | Claude Sonnet 4.5 | 85.12 |
| 3 | GPT-5.1 (Medium) | 85.12 |
| 4 | GPT-5.2 (Medium) | 84.11 |
| 5 | GLM-4.7 FP8 | 84.03 |
| 6 | Qwen 3 235B A22B (Thinking) | 74.84 |
| 7 | GPT-OSS-20B (High) | 69.86 |
| 8 | GPT-OSS-120B (High) | 68.15 |
| 9 | Qwen 3 30B A3B (Thinking) | 62.77 |
| 10 | Qwen 3 8B (Thinking) | 55.61 |
Interactive version: theaggregate.ai/benchmark?slug=medmarks-agentclinic · How the rankings work · Data refreshed daily, snapshot 2026-07-22.