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 | 0.8846 | $214.03 | 690.9M | 2.9M | 15h 36m | 1506 | 34 |
| 2 | Claude Opus 5.5 | 0.8817 | $132.91 | 426.3M | 1.8M | 6h 35m | 922 | 25 |
| 3 | Claude Opus 5.5 | 0.8769 | $159.17 | 476.8M | 2.5M | 11h 3m | 1094 | 29 |
| 4 | Claude Opus 5.5 | 0.7510 | $78.62 | 226.6M | 1.3M | 6h 44m | 478 | 12 |
| 5 | Claude Opus 5.5 | 0.5826 | $111.15 | 316.0M | 1.9M | 10h 19m | 731 | 29 |