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 | 0.5250 | $5.04 | 80.8M | 333k | 14h 2m | 284 | 2 |
| 2 | GLM-5.3 Flash | 0.2214 | $3.70 | 65.2M | 342k | 14h 4m | 250 | 2 |
| 3 | GLM-5.3 Flash | 0.1214 | $5.52 | 97.5M | 412k | 13h 40m | 313 | 9 |
| 4 | GLM-5.3 Flash | 0.0500 | $3.94 | 72.4M | 357k | 14h 22m | 251 | 2 |
| 5 | GLM-5.3 Flash | 0.0179 | $4.71 | 110.5M | 429k | 14h 10m | 330 | 7 |