–

wapi · No library run 1

29.4Library score
Pass rate61.3%
Simplicity65.6
Problems $$5.68

Implementers get no library and write everything from scratch.

ImplementerPass rateSimplicityCostScore
DeepSeek V4.1 Flash · mini-SWE62.6%65.2$0.91
31.2
GLM 5.3 Flash · mini-SWE54.9%68.7$3.04
25.5
GPT-5.6 Luna · Codex66.4%62.9$1.73
31.4
ProblemDeepSeek V4.1 FlashGLM 5.3 FlashGPT-5.6 LunaMean
agent-runtime-protocol2.62.537.514.2reservation-receipts0.757.20.619.5page-publish-links4.03.716.68.1catalog-api22.826.915.421.7portal-routing67.315.049.243.8index-gateway57.21.77.422.1chain-index53.846.759.253.2chain-archive56.650.137.448.0project-routes12.020.241.224.5callback-registry34.831.049.738.5