wapi · Kimi K3 run 1

33.3Library score
Pass rate66.1%
Simplicity65.7
Problems $$12.60
Library $$7.33
The library
wapi · designed by Kimi K3 (mini-SWE)2,495 source lines
ImplementerPass rateSimplicityCostScore
DeepSeek V4.1 Flash · mini-SWE73.0%67.8$1.94
40.6
GLM 5.3 Flash · mini-SWE57.2%58.3$8.05
22.8
GPT-5.6 Luna · Codex68.0%71.1$2.61
36.6
ProblemDeepSeek V4.1 FlashGLM 5.3 FlashGPT-5.6 LunaMean

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