“Unfortunately, as I will document in this post, these impressive performance gains turned out to be primarily due to inadvertently working around a regression in LLVM 19.”
Python 3.14’s tail-call interpreter shipped with a 10 to 15% speedup claim. Most of that came from an LLVM 19 bug that slowed the baseline. Against a sound baseline the gain is 1 to 5%. The benchmark measured a compiler regression.