Vulnerability Exploitation on CVEBench Zero-Day
18.3pass@1Seed Agent
Evaluation Results
| Method | Links | |||||||||
|---|---|---|---|---|---|---|---|---|---|---|
| Seed AgentBackbone=Kimi-K2.52026.05 | 18.3 | 19.8 | 20 | — | — | — | — | — | — | |
| Seed AgentBackbone=Qwen3-235B2026.05 | 14.7 | 15 | 15 | — | — | — | — | — | — | |
| Seed AgentBackbone=MiniMax-M2.52026.05 | 13.4 | 16.9 | 17.5 | — | — | — | — | — | — | |
| Seed AgentBackbone=DeepSeek-V3.12026.05 | 13 | 14.7 | 15 | — | — | — | — | — | — | |
| ACEBackbone=DeepSeek-V3.12026.05 | — | — | — | 15 | 0 | — | — | — | — | |
| ACEBackbone=Kimi-K2.52026.05 | — | — | — | 20 | 0 | — | — | — | — | |
| ACEBackbone=MiniMax-M2.52026.05 | — | — | — | 12.5 | -5 | — | — | — | — | |
| ACEBackbone=Qwen3-235B2026.05 | — | — | — | 15 | 0 | — | — | — | — | |
| CyAgentBackbone=DeepSeek-V3.12026.05 | — | — | — | — | — | — | — | 15 | — | |
| CyAgentBackbone=Kimi-K2.52026.05 | — | — | — | — | — | — | — | 22.5 | — | |
| CyAgentBackbone=MiniMax-M2.52026.05 | — | — | — | — | — | — | — | 17.5 | — | |
| CyAgentBackbone=Qwen3-235B2026.05 | — | — | — | — | — | — | — | 17.5 | — | |
| CyberEvolverBackbone=DeepSeek-V3.12026.05 | — | — | — | — | — | 25 | 10 | — | — | |
| CyberEvolverBackbone=Kimi-K2.52026.05 | — | — | — | — | — | 37.5 | 17.5 | — | — | |
| CyberEvolverBackbone=MiniMax-M2.52026.05 | — | — | — | — | — | 27.5 | 10 | — | — | |
| CyberEvolverBackbone=Qwen3-235B2026.05 | — | — | — | — | — | 32.5 | 17.5 | — | — | |
| T-AgentBackbone=DeepSeek-V3.12026.05 | — | — | — | — | — | — | — | — | 17.5 | |
| T-AgentBackbone=Kimi-K2.52026.05 | — | — | — | — | — | — | — | — | 35 | |
| T-AgentBackbone=MiniMax-M2.52026.05 | — | — | — | — | — | — | — | — | 20 | |
| T-AgentBackbone=Qwen3-235B2026.05 | — | — | — | — | — | — | — | — | 22.5 |