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 | 0.3059 | $79.30 | 127.1M | 413k | 18h 52m | 388 | 8 |
| 2 | Claude Opus 5 | 0.1513 | $87.00 | 142.2M | 403k | 18h 34m | 374 | 4 |
| 3 | Claude Opus 5 | 0.1073 | $79.88 | 127.0M | 429k | 18h 13m | 353 | 6 |
| 4 | Claude Opus 5 | 0.0711 | $105.95 | 145.9M | 442k | 17h 1m | 380 | 4 |
| 5 | Claude Opus 5 | 0.0519 | $94.80 | 144.0M | 475k | 17h 47m | 401 | 7 |