uglypie · Kimi K3 run 1

54.1Library score
Pass rate87.9%
Simplicity67.4
Problems $$4.46
Library $$7.64
The library
uglypie · designed by Kimi K3 (mini-SWE)4,570 source lines
ImplementerPass rateSimplicityCostScore
DeepSeek V4.1 Flash · mini-SWE90.9%67.2$1.36
56.5
GLM 5.3 Flash · mini-SWE88.9%70.8$2.01
56.3
GPT-5.6 Luna · Codex84.0%64.1$1.08
49.4
ProblemDeepSeek V4.1 FlashGLM 5.3 FlashGPT-5.6 LunaMean

Click a problem to see how each solution's score is built.