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 | 0.3530 | $39.56 | 333.6M | 688k | 9h 57m | 942 | 3 |
| 2 | Gemini 3.8 Flash | 0.3004 | $49.60 | 408.7M | 919k | 17h 49m | 1099 | 1 |
| 3 | Gemini 3.8 Flash | 0.2989 | $50.71 | 448.6M | 743k | 11h 49m | 1104 | 1 |
| 4 | Gemini 3.8 Flash | 0.0317 | $65.98 | 570.2M | 1.4M | 17h 58m | 1492 | 1 |
| 5 | Gemini 3.8 Flash | 0.0309 | $52.78 | 424.0M | 773k | 11h 32m | 1103 | 2 |