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 | 0.9207 | $412.86 | 201.8M | 473k | 7h 46m | 595 | 12 |
| 2 | GPT-6 Astra | 0.9186 | $1381.68 | 662.3M | 1.4M | 17h 52m | 1621 | 26 |
| 3 | GPT-6 Astra | 0.9168 | $2130.53 | 1024.2M | 1.9M | 19h 50m | 2384 | 45 |
| 4 | GPT-6 Astra | 0.9131 | $2239.25 | 1067.1M | 2.1M | 19h 57m | 2429 | 54 |
| 5 | GPT-6 Astra | 0.8800 | $604.79 | 292.5M | 576k | 8h 29m | 715 | 11 |