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 | 0.9930 | $40.59 | 65.5M | 322k | 8h 40m | 241 | 9 |
| 2 | Claude Fable 5.1 | 0.9895 | $33.56 | 52.1M | 309k | 8h 2m | 221 | 8 |
| 3 | Claude Fable 5.1 | 0.3345 | $30.45 | 45.8M | 311k | 4h 14m | 174 | 15 |
| 4 | Claude Fable 5.1 | 0.2017 | $30.51 | 47.2M | 284k | 6h 48m | 194 | 9 |
| 5 | Claude Fable 5.1 | 0.0778 | $29.88 | 44.4M | 316k | 2h 48m | 183 | 13 |