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 | 0.0190 | $30.46 | 330.6M | 252k | 3h 24m | 709 | 11 |
| 2 | Gemini 3.8 Flash | 0.0058 | $23.48 | 261.8M | 192k | 2h 8m | 563 | 4 |
| 3 | Gemini 3.8 Flash | 0.0021 | $39.53 | 434.9M | 429k | 7h 2m | 1119 | 13 |
| 4 | Gemini 3.8 Flash | 0.0000 | $24.27 | 258.5M | 198k | 4h 49m | 607 | 3 |
| 5 | Gemini 3.8 Flash | 0.0000 | $21.47 | 225.2M | 181k | 2h 13m | 571 | 5 |