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 | 0.9214 | $95.87 | 150.8M | 548k | 9h 35m | 337 | 5 |
| 2 | Claude Opus 5 | 0.9179 | $89.22 | 139.0M | 522k | 13h 10m | 320 | 6 |
| 3 | Claude Opus 5 | 0.8866 | $79.79 | 122.4M | 498k | 14h 37m | 303 | 4 |
| 4 | Claude Opus 5 | 0.7648 | $96.45 | 153.7M | 524k | 15h 7m | 358 | 5 |
| 5 | Claude Opus 5 | 0.6837 | $80.75 | 120.3M | 550k | 6h 30m | 272 | 7 |