List your company

Long-horizon cascade synthesis on PBEBench Hard

85.8AccuracyBoK + All Symbolic
22.46438.90755.3571.793May 6, 2026

Evaluation results

MethodLinks
BoK + All SymbolicSystem Type=Hybrid (symbolic + LLM fallback), Tokens (M)=71.6, Cost ($)=19.892026.0585.80.95792.74.64
All SymbolicSystem Type=Symbolic solvers, Tokens (M)=0, Cost ($)=7.112026.0584.70.99298.64.56
BoK + CC + QOSystem Type=Hybrid (symbolic + LLM fallback), Tokens (M)=87.6, Cost ($)=18.482026.0583.50.953191.95.48
CC + QOSystem Type=Symbolic solvers, Tokens (M)=0, Cost ($)=2.852026.0581.20.990598.35.35
BoK + QOSystem Type=Hybrid (symbolic + LLM fallback), Tokens (M)=114.5, Cost ($)=21.262026.0580.70.949691.35.48
BoK + CCSystem Type=Hybrid (symbolic + LLM fallback), Tokens (M)=130.0, Cost ($)=25.082026.0579.40.950891.47.82
QO SolverSystem Type=Symbolic solvers, Tokens (M)=0, Cost ($)=0.852026.0574.70.983696.85.26
CC SolverSystem Type=Symbolic solvers, Tokens (M)=0, Cost ($)=2.002026.0569.70.987397.28.06
BoKSystem Type=LLM-only methods, Tokens (M)=332.1, Cost ($)=57.832026.0568.40.942889.95.14
Qwen2.5-32B-A14B (OpenHands)System Type=LLM-only methods, Tokens (M)=1046.8, Cost ($)=∼2242026.0524.90.841173.21.36