P

roadkill · Production library run 2

25.7Library score
Pass rate67.0%
Simplicity44.8
Problems $$11.96

Implementers use the real production library (uxsim).

ImplementerPass rateSimplicityCostScore
DeepSeek V4.1 Flash · mini-SWE76.7%47.3$1.67
30.0
GLM 5.3 Flash · mini-SWE65.4%39.7$7.98
22.1
GPT-5.6 Luna · Codex58.8%47.5$2.32
24.9
ProblemDeepSeek V4.1 FlashGLM 5.3 FlashGPT-5.6 LunaMean
transit-route-check35.622.03.220.3live-session7.26.940.018.0charging-routes50.31.60.017.3link-time-bins83.477.242.567.7explicit-fleet51.30.045.332.2braess-network10.128.536.825.1signal-plan-sweep7.812.055.425.1merge-spillback34.70.00.011.6incident-guidance14.60.53.56.2corridor-replications0.035.60.011.9speed-limit-control29.916.434.126.8reactive-signals19.142.672.544.7signal-progression30.926.511.523.0gridlock-gating41.642.338.140.7workzone-recovery50.928.42.327.2district-los12.613.913.913.4