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 Fable 5.1 | 0.5126 | $87.17 | 154.8M | 404k | 6h 52m | 408 | 12 |
| 2 | Claude Fable 5.1 | 0.4984 | $96.24 | 173.0M | 403k | 7h 34m | 486 | 21 |
| 3 | Claude Fable 5.1 | 0.4822 | $86.06 | 152.3M | 409k | 8h 44m | 402 | 23 |
| 4 | Claude Fable 5.1 | 0.4767 | $88.31 | 156.7M | 413k | 3h 39m | 397 | 10 |
| 5 | Claude Fable 5.1 | 0.3738 | $86.02 | 153.2M | 388k | 5h 28m | 424 | 9 |