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 | 0.4639 | $1.55 | 83.0M | 253k | 19h 5m | 403 | 5 |
| 2 | DeepSeek V4 Flash Vision Exp | 0.0000 | $1.66 | 90.1M | 258k | 19h 29m | 392 | 3 |
| 3 | DeepSeek V4 Flash Vision Exp | 0.0000 | $1.99 | 110.4M | 289k | 18h 13m | 487 | 4 |
| 4 | DeepSeek V4 Flash Vision Exp | 0.0000 | $1.55 | 84.2M | 243k | 18h 42m | 398 | 5 |
| 5 | DeepSeek V4 Flash Vision Exp | 0.0000 | $1.68 | 93.8M | 242k | 19h 27m | 452 | 2 |