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 | 0.6886 | $9.16 | 61.0M | 246k | 3h 25m | 267 | 8 |
| 2 | Gemini 3.8 Flash | 0.6750 | $4.96 | 25.6M | 195k | 1h 47m | 172 | 5 |
| 3 | Gemini 3.8 Flash | 0.6169 | $10.88 | 84.0M | 323k | 4h 38m | 300 | 2 |
| 4 | Gemini 3.8 Flash | 0.6021 | $4.98 | 34.7M | 202k | 2h 25m | 218 | 4 |
| 5 | Gemini 3.8 Flash | 0.5930 | $6.20 | 47.0M | 194k | 4h 21m | 224 | 4 |