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 | 0.6474 | $297.67 | 145.6M | 335k | 19h 47m | 441 | 14 |
| 2 | GPT-6 Astra | 0.6387 | $27.55 | 17.3M | 152k | 18h 9m | 135 | 12 |
| 3 | GPT-6 Astra | 0.6362 | $98.77 | 53.1M | 216k | 17h 56m | 257 | 14 |
| 4 | GPT-6 Astra | 0.6334 | $244.67 | 123.1M | 350k | 18h 49m | 445 | 13 |
| 5 | GPT-6 Astra | 0.0000 | $488.12 | 238.0M | 474k | 17h 46m | 610 | 16 |