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 | 0.0000 | $90.95 | 821.2M | 1.4M | 20h 0m | 2038 | 1 |
| 2 | Gemini 3.8 Flash | 0.0000 | $69.03 | 618.4M | 870k | 14h 14m | 1477 | 3 |
| 3 | Gemini 3.8 Flash | 0.0000 | $77.67 | 724.4M | 1.0M | 15h 44m | 1676 | 2 |
| 4 | Gemini 3.8 Flash | 0.0000 | $82.03 | 739.4M | 1.1M | 17h 33m | 1797 | 1 |
| 5 | Gemini 3.8 Flash | 0.0000 | $66.46 | 629.7M | 880k | 14h 44m | 1599 | 3 |