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 | 0.5599 | $663.30 | 1199.7M | 2.6M | 18h 57m | 2674 | 17 |
| 2 | Claude Fable 5.1 | 0.5146 | $566.08 | 1019.0M | 2.3M | 19h 11m | 2360 | 13 |
| 3 | Claude Fable 5.1 | 0.5049 | $432.88 | 762.8M | 2.1M | 18h 58m | 1732 | 13 |
| 4 | Claude Fable 5.1 | 0.4886 | $363.29 | 631.2M | 2.0M | 18h 58m | 1562 | 11 |
| 5 | Claude Fable 5.1 | 0.4844 | $390.64 | 686.7M | 2.0M | 19h 13m | 1611 | 26 |