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 | GLM-5.3 Flash | 0.5051 | $24.75 | 719.3M | 1.5M | 17h 38m | 1725 | 1 |
| 2 | GLM-5.3 Flash | 0.4646 | $19.29 | 592.9M | 1.1M | 15h 32m | 1400 | 4 |
| 3 | GLM-5.3 Flash | 0.3737 | $22.66 | 655.9M | 1.7M | 20h 0m | 1644 | 1 |
| 4 | GLM-5.3 Flash | 0.3737 | $15.05 | 404.9M | 975k | 14h 18m | 975 | 2 |
| 5 | GLM-5.3 Flash | 0.3434 | $27.30 | 749.1M | 1.3M | 19h 17m | 1599 | 3 |