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 | GPT-6 Astra | 0.9091 | $2887.40 | 1385.4M | 2.5M | 19h 50m | 3121 | 180 |
| 2 | GPT-6 Astra | 0.8990 | $347.32 | 164.1M | 405k | 2h 56m | 414 | 7 |
| 3 | GPT-6 Astra | 0.8990 | $599.18 | 284.8M | 531k | 4h 2m | 616 | 7 |
| 4 | GPT-6 Astra | 0.8889 | $677.88 | 324.1M | 558k | 4h 32m | 694 | 7 |
| 5 | GPT-6 Astra | 0.8687 | $172.72 | 83.5M | 308k | 2h 47m | 288 | 4 |