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 | 0.0110 | $29.12 | 1761.2M | 2.9M | 19h 19m | 3911 | 24 |
| 2 | DeepSeek V4 Flash Vision Exp | 0.0053 | $22.61 | 1343.5M | 2.5M | 19h 20m | 2867 | 26 |
| 3 | DeepSeek V4 Flash Vision Exp | 0.0049 | $20.65 | 1205.0M | 2.5M | 19h 56m | 2498 | 19 |
| 4 | DeepSeek V4 Flash Vision Exp | 0.0049 | $17.94 | 1097.8M | 1.6M | 18h 58m | 2348 | 14 |
| 5 | DeepSeek V4 Flash Vision Exp | 0.0049 | $20.78 | 1251.0M | 2.1M | 19h 4m | 2804 | 15 |