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 | DeepSeek V4 Flash Vision Exp | 0.1786 | $2.02 | 107.0M | 337k | 14h 34m | 328 | 2 |
| 2 | DeepSeek V4 Flash Vision Exp | 0.1214 | $2.99 | 164.7M | 448k | 8h 44m | 423 | 5 |
| 3 | DeepSeek V4 Flash Vision Exp | 0.0357 | $2.71 | 145.6M | 448k | 14h 17m | 377 | 7 |
| 4 | DeepSeek V4 Flash Vision Exp | 0.0143 | $2.49 | 130.6M | 440k | 11h 4m | 400 | 10 |
| 5 | DeepSeek V4 Flash Vision Exp | 0.0036 | $2.74 | 145.9M | 463k | 9h 28m | 372 | 6 |