weft · Opus 5 run 3
27.5Run score
Pass rate48.6%
Simplicity57.9
Problems $$26.14
Opus 5 uses the real production library (chumsky).
ImplementerPass rateSimplicityCostScore
Opus 5 · mini-SWE48.6%57.9$26.14
ProblemOpus 5
Click a problem to see how each solution's score is built.