Capture The Flag on 30 CTF Challenges
0.6333Success Rate (30 Challenges)Executor
Evaluation Results
| Method | Links | ||||
|---|---|---|---|---|---|
| ExecutorModel=GPT-52026.04 | 0.6333 | 31.56 | 0.9 | 336.74 | |
| Executor + EvaluatorModel=GPT-52026.04 | 0.6333 | 28.76 | 0.64 | 799.3 | |
| Planner + Executor + EvaluatorModel=GPT-52026.04 | 0.6333 | 24.09 | 0.59 | 925.8 | |
| claude-codeModel=Opus-4-52026.04 | 0.6333 | 45.52 | 1.26 | 284.72 | |
| ExecutorModel=GPT-4.12026.04 | 0.3 | 36.74 | 1.43 | 162.13 |