I was under the impression Python was magnitudes slower than compiled native languages, even when you aren't abusing Pythonic features that are computationally

I was under the impression Python was magnitudes slower than compiled native languages, even when you aren't abusing Pythonic features that are computationally expensive like list comprehensions. And this isn't even in Pypy where this loop would be JIT'd and then run natively anyway, unless cpython does some kind of JIT? Is is just that the buffer is so large that the kernel handler is taking so much of a % of the runtime that the inefficiencies of the language stop mattering?