weft · Grok 4.6 run 1
34.6Library score
Pass rate81.4%
Simplicity50.4
Problems $$11.14
Library $$5.91
The library
weft · designed by Grok 4.6 (mini-SWE)5,200 source lines
ImplementerPass rateSimplicityCostScore
DeepSeek V4.1 Flash · mini-SWE85.6%49.7$3.66
GLM 5.3 Flash · mini-SWE77.8%53.0$5.14
GPT-5.6 Luna · Codex80.7%48.3$2.33
ProblemDeepSeek V4.1 FlashGLM 5.3 FlashGPT-5.6 LunaMean
bag-rules34.320.326.827.1ninja-build33.128.332.731.4mail-addresses41.136.954.144.1property-paths45.136.841.141.0formula-ledger32.348.728.336.4record-spans68.473.846.663.0block-events48.653.642.648.3command-errors24.126.921.724.2route-tags51.94.728.328.3template-audit33.928.219.427.2grammar-inventory28.142.524.831.8grammar-trace24.749.944.939.8nesting-guard18.214.211.514.6unicode-spans27.721.831.427.0