plaid · Kimi K3 run 2

40.4Library score
Pass rate93.3%
Simplicity50.7
Problems $$10.97
Library $$5.17
The library
plaid · designed by Kimi K3 (mini-SWE)1,697 source lines
ImplementerPass rateSimplicityCostScore
DeepSeek V4.1 Flash · mini-SWE98.8%49.9$2.61
48.8
GLM 5.3 Flash · mini-SWE90.3%51.6$5.83
28.7
GPT-5.6 Luna · Codex89.7%50.9$2.53
43.5
ProblemDeepSeek V4.1 FlashGLM 5.3 FlashGPT-5.6 LunaMean
ledger-balance51.515.642.036.3typed-settings46.540.665.850.9keymap-resolver48.449.546.548.1template-trace49.50.049.232.9recipe-graph53.330.044.242.5scoped-eval44.80.034.326.4temporal-normalizer47.50.038.228.6script-recovery49.30.00.016.4stack-audit39.338.743.840.6query-resolution62.667.757.662.6selection-flattener57.148.149.451.5schema-resolver53.545.952.050.5grammar-crossref52.754.353.053.3config-digest41.40.031.124.2shape-atlas34.640.846.140.5