SnorkelLegal pass@5
Professional Work
Published pass@5 on the 200-task SnorkelLegal release; a correlated diagnostic of the pass@1 leaderboard.
Top models (higher is better)
| Model | Score |
|---|---|
| Kimi K3 | 65.3 |
| Grok 4.6 | 64.6 |
| GLM-5.3 | 63.6 |
| Claude Fable 5.1 | 60.5 |
| Muse Spark 1.3 | 55.5 |
| Grok 4.7 | 53.4 |
| Claude Opus 5.5 | 51.5 |
| Claude Opus 5 | 50.0 |
| Gemini 3.8 Flash | 42.0 |
| GPT-6 Astra | 35.1 |