simu · Kimi K3 run 1

40.2Library score
Pass rate84.9%
Simplicity55.1
Problems $$5.42
Library $$7.07
The library
simu · designed by Kimi K3 (mini-SWE)1,158 source lines
ImplementerPass rateSimplicityCostScore
DeepSeek V4.1 Flash · mini-SWE91.3%53.8$1.22
46.4
GLM 5.3 Flash · mini-SWE87.1%60.8$2.92
42.1
GPT-5.6 Luna · Codex76.4%51.2$1.27
32.2
ProblemDeepSeek V4.1 FlashGLM 5.3 FlashGPT-5.6 LunaMean

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