Medmarks - HealthBench Consensus — leaderboard
Metric: Score (%). Source: medmarks.ai. 12 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | GPT-5.2 (Medium) | 66.16 |
| 2 | GPT-5.1 (Medium) | 65.72 |
| 3 | GPT-OSS-120B (High) | 60.45 |
| 4 | Qwen 3 235B A22B (Thinking) | 58.78 |
| 5 | Gemini 3 Pro (Preview) | 54.43 |
| 6 | GLM-4.7 FP8 | 52.78 |
| 7 | Qwen 3 30B A3B (Thinking) | 52.47 |
| 8 | Claude Sonnet 4.5 | 51.84 |
| 9 | Qwen 3 8B (Thinking) | 49.06 |
| 10 | GPT-OSS-20B (High) | 48.83 |
Interactive version: theaggregate.ai/benchmark?slug=medmarks-healthbench-consensus · How the rankings work · Data refreshed daily, snapshot 2026-07-22.