Medmarks - SuperGPQA Medicine Easy: leaderboard
Metric: Score (%). Source: medmarks.ai. 71 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Gemini 3 Pro (Preview) | 76.86 |
| 2 | GPT-5.1 (Medium) | 68.87 |
| 3 | GPT-5.2 (Medium) | 67.55 |
| 4 | Claude Sonnet 4.5 | 67.51 |
| 5 | Qwen 3 235B A22B (Thinking) | 66.7 |
| 6 | Grok 4 | 66.52 |
| 7 | GLM-4.7 FP8 | 66.12 |
| 8 | Qwen 3 Next 80B A3B Instruct | 62.67 |
| 9 | Qwen 3 Next 80B A3B (Thinking) | 62.63 |
| 10 | MiniMax-M2.1 | 59.19 |
| 11 | Qwen 3 VL 30B A3B (Thinking) | 58.82 |
| 12 | GLM 4.5 Air | 58.23 |
| 13 | INTELLECT-3 | 56.69 |
| 14 | MiniMax-M2 | 55.45 |
| 15 | Ling-flash-2.0 | 54.68 |
Interactive version: theaggregate.ai/benchmark?slug=medmarks-supergpqa-medicine-easy · How It Works · Data refreshed daily, snapshot 2026-09-05.