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 3.7 Flash | 0.6517 | $199.28 | 2123.1M | 2.6M | 19h 55m | 4917 | 62 |
| 2 | Gemini 3.7 Flash | 0.5376 | $121.24 | 1110.9M | 1.5M | 19h 27m | 2524 | 37 |
| 3 | Gemini 3.7 Flash | 0.4365 | $75.56 | 629.3M | 934k | 19h 55m | 1331 | 3 |
| 4 | Gemini 3.7 Flash | 0.4242 | $33.92 | 324.0M | 891k | 3h 25m | 780 | 8 |
| 5 | Gemini 3.7 Flash | 0.3137 | $25.24 | 282.5M | 578k | 1h 28m | 590 | 2 |