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 4 Carbon | 0.9611 | $66.56 | 305.1M | 1.5M | 9h 30m | 743 | 19 |
| 2 | Gemini 4 Carbon | 0.9539 | $71.82 | 359.3M | 1.4M | 11h 9m | 825 | 21 |
| 3 | Gemini 4 Carbon | 0.9419 | $55.72 | 253.6M | 1.2M | 7h 41m | 602 | 18 |
| 4 | Gemini 4 Carbon | 0.9296 | $53.84 | 256.6M | 1.2M | 7h 35m | 618 | 14 |
| 5 | Gemini 4 Carbon | 0.9267 | $34.26 | 184.5M | 928k | 5h 56m | 437 | 13 |