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 Opus 5 | 0.8211 | $159.35 | 224.3M | 1.3M | 7h 45m | 513 | 7 |
| 2 | Claude Opus 5 | 0.8211 | $245.12 | 359.5M | 1.8M | 15h 10m | 826 | 19 |
| 3 | Claude Opus 5 | 0.8130 | $286.81 | 434.1M | 1.9M | 15h 11m | 970 | 12 |
| 4 | Claude Opus 5 | 0.8130 | $184.58 | 276.3M | 1.3M | 15h 20m | 620 | 7 |
| 5 | Claude Opus 5 | 0.7886 | $115.90 | 152.7M | 1.1M | 6h 12m | 390 | 8 |