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 | 0.5042 | $12.98 | 108.1M | 394k | 2h 18m | 329 | 5 |
| 2 | Gemini 3.8 Flash | 0.4012 | $12.80 | 133.1M | 348k | 1h 19m | 360 | 2 |
| 3 | Gemini 3.8 Flash | 0.3668 | $11.65 | 115.4M | 351k | 1h 8m | 353 | 6 |
| 4 | Gemini 3.8 Flash | 0.2792 | $7.71 | 71.8M | 268k | 1h 9m | 283 | 2 |
| 5 | Gemini 3.8 Flash | 0.0000 | $15.57 | 160.4M | 440k | 1h 35m | 409 | 1 |