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 | 0.5259 | $112.64 | 181.6M | 606k | 17h 57m | 454 | 17 |
| 2 | Claude Opus 5 | 0.4532 | $73.01 | 109.0M | 516k | 14h 40m | 352 | 12 |
| 3 | Claude Opus 5 | 0.4439 | $118.20 | 194.6M | 584k | 18h 55m | 507 | 14 |
| 4 | Claude Opus 5 | 0.4351 | $112.61 | 183.8M | 577k | 18h 34m | 471 | 19 |
| 5 | Claude Opus 5 | 0.4296 | $84.72 | 133.1M | 506k | 18h 0m | 383 | 12 |