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 | 0.8586 | $166.41 | 276.3M | 1.2M | 9h 57m | 632 | 17 |
| 2 | Claude Fable 5.1 | 0.7677 | $93.05 | 145.4M | 837k | 3h 54m | 417 | 4 |
| 3 | Claude Fable 5.1 | 0.7071 | $96.79 | 153.4M | 827k | 4h 2m | 413 | 8 |
| 4 | Claude Fable 5.1 | 0.6970 | $69.08 | 107.8M | 623k | 2h 37m | 227 | 2 |
| 5 | Claude Fable 5.1 | 0.6667 | $96.05 | 148.5M | 898k | 4h 16m | 406 | 6 |