heretical · Opus 5.5 run 1

46.6Library score
Pass rate87.5%
Simplicity60.3
Problems $$11.37
Library $$8.57
The library
heretical · designed by Opus 5.5 (mini-SWE)3,501 source lines
ImplementerPass rateSimplicityCostScore
DeepSeek V4.1 Flash · mini-SWE89.5%62.5$4.14
50.2
GLM 5.3 Flash · mini-SWE84.2%66.1$4.70
47.8
GPT-5.6 Luna · Codex88.7%52.2$2.53
41.9
ProblemDeepSeek V4.1 FlashGLM 5.3 FlashGPT-5.6 LunaMean

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