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 | 0.9533 | $570.71 | 935.1M | 2.7M | 19h 25m | 2060 | 23 |
| 2 | Claude Opus 5 | 0.8947 | $669.61 | 1119.9M | 2.9M | 19h 5m | 2462 | 32 |
| 3 | Claude Opus 5 | 0.8643 | $526.73 | 857.9M | 2.6M | 14h 11m | 1791 | 32 |
| 4 | Claude Opus 5 | 0.8500 | $580.43 | 975.9M | 2.5M | 18h 59m | 2103 | 31 |
| 5 | Claude Opus 5 | 0.8367 | $450.82 | 756.2M | 1.9M | 15h 49m | 1560 | 11 |