Benchmark scores get restated when harnesses change or models are re-run. This page keeps the trail instead of quietly overwriting it.
No historical snapshots recorded yet — the current scores are shown below.
| # | Model | Score | Observed |
|---|---|---|---|
| 1 | Claude Opus 5 | 43.3% | 2026-10-19 |
| 2 | GPT 5.6-Sol | 34.4% | 2026-10-19 |
| 3 | DeepSeek V4.1-Flash 0910 | 30% | 2026-10-19 |
| 4 | GLM 5.3 | 28.3% | 2026-10-19 |
| 5 | Kimi K3 | 17.7% | 2026-10-19 |
| 6 | DeepSeek V4-Pro 0813 | 11.8% | 2026-10-19 |
| 7 | DeepSeek V4-Flash 0731 | 7.6% | 2026-10-19 |
Nothing has been overwritten yet.