Verify no false positives on clean Sozu load balancer binary (no backdoor inserted).
Performance
| Model | Pass Rate | Runs | Avg Cost | Avg Time |
|---|---|---|---|---|
| deepseek-v3.2 | 100% | | $0.03 | 7m |
| gpt-5.2-codex | 100% | | $0.05 | 1m |
| gpt-5 | 100% | | $0.07 | 2m |
| claude-haiku-4.5 | 100% | | $0.11 | 1m |
| glm-4.7 | 100% | | $0.13 | 18m |
| gemini-3-flash-preview | 100% | | $0.13 | 3m |
| gpt-5.3-codex | 100% | | $0.14 | 2m |
| gpt-5.3-codex-high | 100% | | $0.20 | 4m |
| grok-4 | 100% | | $0.22 | 5m |
| gemini-2.5-pro | 100% | | $0.25 | 5m |
| claude-sonnet-4.5 | 100% | | $0.39 | 4m |
| gemini-3-pro-preview | 100% | | $0.43 | 5m |
| gpt-5.2-codex-high | 100% | | $0.53 | 7m |
| gemini-3.1-pro-preview | 100% | | $0.54 | 9m |
| gpt-5.2-codex-xhigh | 100% | | $0.96 | 19m |
| claude-opus-4.6 | 100% | | $3.98 | 43m |
| glm-5 | 67% | | $0.30 | 17m |
| gpt-5.3-codex-xhigh | 67% | | $0.49 | 47m |
| claude-opus-4.5 | 33% | | $1.57 | 15m |
| kimi-k2.5 | 0% | | $0.14 | 10m |
| claude-sonnet-4 | 0% | | $0.40 | 4m |
All product names, logos, and brands (™/®) are the property of their respective owners; they're used here solely for identification and comparison, and their use does not imply affiliation, endorsement, or sponsorship.