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 | 0.9212 | $529.35 | 935.4M | 1.6M | 17h 43m | 2148 | 24 |
| 2 | Claude Opus 5 | 0.9178 | $356.58 | 629.0M | 1.1M | 18h 31m | 1391 | 16 |
| 3 | Claude Opus 5 | 0.9127 | $588.34 | 1052.2M | 1.6M | 19h 25m | 2364 | 25 |
| 4 | Claude Opus 5 | 0.9110 | $505.22 | 890.6M | 1.6M | 16h 8m | 2006 | 21 |
| 5 | Claude Opus 5 | 0.8921 | $377.04 | 660.7M | 1.2M | 15h 32m | 1403 | 21 |