Every run: the agent's full trajectory, its submissions over the 20-hour budget, and the final score.
| # | Model | Reward | Cost | In tok | Out tok | Time | Steps | Submits |
|---|---|---|---|---|---|---|---|---|
| 1 | Claude Opus 5.5 | 0.9894 | $168.41 | 622.4M | 1.6M | 10h 59m | 1389 | 39 |
| 2 | Claude Opus 5.5 | 0.9885 | $192.52 | 720.9M | 1.8M | 13h 2m | 1670 | 53 |
| 3 | Claude Opus 5.5 | 0.9853 | $159.11 | 577.5M | 1.6M | 9h 35m | 1338 | 18 |
| 4 | Claude Opus 5.5 | 0.9829 | $190.35 | 691.9M | 2.0M | 9h 34m | 1602 | 24 |
| 5 | Claude Opus 5.5 | 0.9733 | $318.71 | 1272.8M | 2.4M | 16h 51m | 2868 | 99 |