List your company

Multi-Agent Collaboration on MABENCH (benign)

93.3BSR Finance COGPT-5.5
79.46883.05986.6590.241Jun 11, 2026

Evaluation results

MethodLinks
GPT-5.5MAS Model=GPT-5.52026.0693.395709010010010076.755450017.564.8
Gemini 3.1 ProMAS Model=Gemini 3.1 Pro2026.0683.382.588.382.510010010063.39027.52.5754572.3
Claude Opus 4.7MAS Model=Claude Opus 4.72026.06809590851001001009010010050569.6