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 | 0.0069 | $10.96 | 663.9M | 1.1M | 19h 2m | 1527 | 2 |
| 2 | DeepSeek V4 Flash Vision Exp | 0.0065 | $10.60 | 633.4M | 1.1M | 15h 56m | 1516 | 8 |
| 3 | DeepSeek V4 Flash Vision Exp | 0.0028 | $12.40 | 756.6M | 1.2M | 19h 8m | 1756 | 39 |
| 4 | DeepSeek V4 Flash Vision Exp | 0.0027 | $6.77 | 415.2M | 638k | 18h 28m | 837 | 7 |
| 5 | DeepSeek V4 Flash Vision Exp | 0.0000 | $11.01 | 666.7M | 1.1M | 19h 11m | 1467 | 11 |