–
uglypie · No library run 2
31.7Library score
Pass rate88.5%
Simplicity39.0
Problems $$3.39
Implementers get no library and write everything from scratch.
ImplementerPass rateSimplicityCostScore
DeepSeek V4.1 Flash · mini-SWE83.5%37.0$0.88
GLM 5.3 Flash · mini-SWE91.0%40.6$1.81
GPT-5.6 Luna · Codex90.9%39.5$0.70
ProblemDeepSeek V4.1 FlashGLM 5.3 FlashGPT-5.6 LunaMean
portal-ingest34.544.241.340.0kml-schema-import17.033.432.327.5news-crawl23.918.122.721.6access-tree23.320.625.923.3page-cleaner41.121.246.936.4exhibit-route19.726.426.924.3hearing-replay34.745.228.536.1festival-revisions41.844.230.638.9page-notes0.040.724.321.6genome-catalog52.146.844.347.7article-rewrite30.029.334.131.1