List your company

Vulnerability Detection on Security Review Benchmark 100-sample Baseline (C0 + SP0)

96.1C0 Detection RatePerplexity Sonar Pro
51.27662.91374.5586.187Feb 18, 2026

Evaluation results

MethodLinks
Perplexity Sonar ProClass=Commercial2026.0296.1
Claude Opus 4.6Class=Commercial2026.0295
GPT-5.2Class=Commercial2026.0293
Gemini 2.5 ProClass=Commercial2026.0291.9
DeepSeek ChatClass=Commercial2026.0289
Llama 3.3 70BClass=Open-Source2026.0272
Qwen 2.5 72BClass=Open-Source2026.0270
DeepSeek-Coder-V2 16BClass=Open-Source2026.0253