simu · DeepSeek V4 Pro run 1

37.0Library score
Pass rate89.0%
Simplicity50.1
Problems $$5.77
Library $$0.28
The library
simu · designed by DeepSeek V4 Pro (mini-SWE)978 source lines
ImplementerPass rateSimplicityCostScore
DeepSeek V4.1 Flash · mini-SWE87.7%47.6$2.92
37.7
GLM 5.3 Flash · mini-SWE91.5%54.0$1.57
32.2
GPT-5.6 Luna · Codex88.6%50.0$1.28
41.3
ProblemDeepSeek V4.1 FlashGLM 5.3 FlashGPT-5.6 LunaMean

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