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 | 0.9974 | $3129.48 | 1515.4M | 2.4M | 19h 55m | 3376 | 49 |
| 2 | GPT-6 Astra | 0.9897 | $2759.22 | 1308.1M | 2.5M | 19h 59m | 2942 | 38 |
| 3 | GPT-6 Astra | 0.9861 | $2721.06 | 1296.9M | 2.6M | 19h 56m | 2978 | 26 |
| 4 | GPT-6 Astra | 0.9793 | $2755.43 | 1303.4M | 2.5M | 19h 59m | 2916 | 37 |
| 5 | GPT-6 Astra | 0.9655 | $2773.42 | 1319.2M | 2.5M | 19h 57m | 2941 | 30 |