AA GPQA Diamond: leaderboard
One item is a graduate-level multiple-choice question in biology, physics or chemistry with four options, taken from the 198-question Diamond subset of GPQA. Artificial Analysis runs the set itself, zero-shot with an instruction prompt and no worked examples, pulls the chosen letter from the reply by regex and scores it pass@1. The number is the percent answered correctly, and higher is better.
Metric: Accuracy (%). Source: artificialanalysis.ai. Saturation forecast: Estimated already saturated. 617 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | GPT-6 Astra (xHigh) | 96.26 | #6 (GPT-6 Astra) |
| 2 | GPT-6 Astra (Max) | 96.06 | #6 (GPT-6 Astra) |
| 3 | Gemini 3.8 Flash (High) | 95.25 | #21 (Gemini 3.8 Flash) |
| 4 | Grok 4.6 (High) | 94.95 | #26 (Grok 4.6) |
| 5 | GPT-6 Astra (High) | 94.95 | #6 (GPT-6 Astra) |
| 6 | Gemini 3.7 Flash (High) | 94.55 | #23 (Gemini 3.7 Flash) |
| 7 | Gemini 3.1 Pro (Preview) | 94.14 | #58 |
| 8 | GPT-5.6 Sol (Max) | 94.14 | #19 (GPT-5.6 Sol) |
| 9 | Muse Spark 1.3 (xHigh) | 94.14 | #24 (Muse Spark 1.3) |
| 10 | GPT-6 Astra (Medium) | 93.94 | #6 (GPT-6 Astra) |
| 11 | Claude Opus 5 (High) | 93.74 | #12 (Claude Opus 5) |
| 12 | Claude Opus 5 (xHigh) | 93.74 | #12 (Claude Opus 5) |
| 13 | GPT-5.5 (xHigh) | 93.54 | #31 (GPT-5.5) |
| 14 | Kimi K3 (Max) | 93.54 | #30 (Kimi K3) |
| 15 | Muse Spark 1.3 (Max) | 93.54 | #24 (Muse Spark 1.3) |
No result here: #3 Claude Opus 5.5, #7 GPT-6.1 Sol, #8 Gemini 4 Argon.
Interactive version: theaggregate.ai/benchmark?slug=aa-gpqa-diamond · How It Works · Data refreshed daily, snapshot 2026-10-20.