List your company

Vulnerability Detection on EVMbench Detect

71.7RecallChaintrix
14.529.3544.259.05May 10, 2026

Evaluation results

MethodLinks
Chaintrix2026.0571.7
Claude Opus 4.6Framework=Claude Code, Agent=best single agent2026.0547.5
Gemini 3.1 ProFramework=OpenCode, Tools=custom tools2026.0537.5
Claude Opus 4.5Framework=OpenCode2026.0535.8
GPT-5.3-CodexVariant=best Codex CLI variant2026.0529.2
Gemini 3 Pro PreviewConfiguration=worst configuration2026.0516.7