–
wapi · No library run 2
33.0Library score
Pass rate62.2%
Simplicity66.9
Problems $$5.39
Implementers get no library and write everything from scratch.
ImplementerPass rateSimplicityCostScore
DeepSeek V4.1 Flash · mini-SWE70.4%64.9$0.81
GLM 5.3 Flash · mini-SWE52.9%68.9$2.82
GPT-5.6 Luna · Codex63.1%66.8$1.76
ProblemDeepSeek V4.1 FlashGLM 5.3 FlashGPT-5.6 LunaMean
agent-runtime-protocol2.028.72.411.0reservation-receipts0.80.70.70.7page-publish-links7.21.51.93.5catalog-api39.034.929.034.3portal-routing62.122.556.847.1index-gateway57.161.869.162.7chain-index59.563.349.557.4chain-archive73.90.415.630.0project-routes38.50.943.327.6callback-registry40.163.663.355.7