List your company

Text-to-SQL on BIRD-147 2023 (test)

97.3SRRAPEX MAS
70.7877.66584.5591.435Apr 2, 2026

Evaluation results

MethodLinks
APEX MASBase Model=GPT-4o, Time (s)=179, Cost ($)=0.362026.0497.399.43952.870.7
APEX MASBase Model=Sonnet, Time (s)=191, Cost ($)=0.452026.0493.998.34760.869.4
MAS CompilerTime (s)=64, Cost ($)=0.402026.0479.693.694.893.767.4
Adaptive SkillTime (s)=55, Cost ($)=0.322026.0477.692.795.393.468
Tools OnlyTime (s)=27, Cost ($)=0.232026.0475.592.594.792.966
Knowledge OnlyTime (s)=12, Cost ($)=0.322026.0474.292.594.993.166.7
Claude Code RawTime (s)=48, Cost ($)=0.332026.0471.891.89090.559.2