Medmarks - PubHealthBench Freeform — leaderboard
Metric: Score (%). Source: medmarks.ai. 12 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | GPT-5.2 (Medium) | 48.09 |
| 2 | GPT-5.1 (Medium) | 46.84 |
| 3 | Gemini 3 Pro (Preview) | 30.13 |
| 4 | Claude Sonnet 4.5 | 24.14 |
| 5 | GLM-4.7 FP8 | 24.14 |
| 6 | GPT-OSS-120B (High) | 21.12 |
| 7 | GPT-OSS-20B (High) | 19.28 |
| 8 | Qwen 3 235B A22B (Thinking) | 13.62 |
| 9 | Qwen 3 30B A3B (Thinking) | 10.53 |
| 10 | Qwen 3 8B (Thinking) | 9.93 |
Interactive version: theaggregate.ai/benchmark?slug=medmarks-pubhealthbench-freeform · How the rankings work · Data refreshed daily, snapshot 2026-07-22.