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 | 0.1100 | $1.27 | 65.0M | 240k | 20h 4m | 306 | 1 |
| 2 | DeepSeek V4 Flash Vision Exp | 0.0860 | $1.47 | 79.6M | 227k | 19h 17m | 390 | 2 |
| 3 | DeepSeek V4 Flash Vision Exp | 0.0160 | $1.33 | 68.9M | 247k | 19h 56m | 345 | 2 |
| 4 | DeepSeek V4 Flash Vision Exp | 0.0160 | $1.13 | 55.9M | 221k | 19h 37m | 264 | 2 |
| 5 | DeepSeek V4 Flash Vision Exp | 0.0000 | $1.61 | 87.5M | 249k | 20h 2m | 395 | 1 |