heretical · GLM 5.3 run 2

33.0Library score
Pass rate82.0%
Simplicity49.4
Problems $$17.77
Library $$16.29
The library
heretical · designed by GLM 5.3 (mini-SWE)5,459 source lines
ImplementerPass rateSimplicityCostScore
DeepSeek V4.1 Flash · mini-SWE86.5%49.1$3.41
36.8
GLM 5.3 Flash · mini-SWE77.3%52.7$10.65
30.9
GPT-5.6 Luna · Codex81.8%46.4$3.72
31.2
ProblemDeepSeek V4.1 FlashGLM 5.3 FlashGPT-5.6 LunaMean

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