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 | 0.5584 | $3.61 | 51.8M | 199k | 18h 29m | 284 | 2 |
| 2 | GLM-5.3 Flash | 0.5292 | $1.71 | 30.3M | 154k | 17h 56m | 222 | 5 |
| 3 | GLM-5.3 Flash | 0.4629 | $3.93 | 71.7M | 248k | 18h 47m | 338 | 2 |
| 4 | GLM-5.3 Flash | 0.3763 | $2.83 | 73.4M | 262k | 16h 13m | 343 | 13 |
| 5 | GLM-5.3 Flash | 0.3496 | $2.98 | 85.3M | 281k | 19h 32m | 394 | 4 |