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 | Gemini 3.8 Flash | 0.4044 | $62.09 | 665.3M | 979k | 9h 34m | 1663 | 2 |
| 2 | Gemini 3.8 Flash | 0.3444 | $76.61 | 826.4M | 1.1M | 12h 14m | 1952 | 4 |
| 3 | Gemini 3.8 Flash | 0.3068 | $59.22 | 604.8M | 923k | 9h 13m | 1507 | 4 |
| 4 | Gemini 3.8 Flash | 0.2640 | $52.50 | 504.3M | 779k | 8h 9m | 1188 | 4 |
| 5 | Gemini 3.8 Flash | 0.1660 | $58.00 | 614.2M | 934k | 9h 4m | 1564 | 2 |