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 | 0.9491 | $569.86 | 1069.9M | 1.4M | 17h 10m | 2285 | 28 |
| 2 | Claude Fable 5.1 | 0.9284 | $552.06 | 1031.1M | 1.5M | 18h 56m | 2091 | 46 |
| 3 | Claude Fable 5.1 | 0.9275 | $456.14 | 848.4M | 1.3M | 18h 57m | 2026 | 39 |
| 4 | Claude Fable 5.1 | 0.8919 | $415.43 | 774.8M | 1.2M | 19h 19m | 1796 | 47 |
| 5 | Claude Fable 5.1 | 0.8712 | $313.43 | 577.0M | 1.0M | 18h 15m | 1283 | 12 |