Python 3.11: an average 25% faster without changing your code

The problem

Python has a long-standing reputation for being slow, and the usual advice was to rewrite hot code in C, or reach for a different tool. For most teams, both are expensive.

What they did

The Faster CPython project set out to speed up the interpreter itself. Per the Python 3.11 release notes, "The main team is funded by Microsoft to work on this full-time. Pablo Galindo Salgado is also funded by Bloomberg LP to work on the project part-time. Finally, many contributors are volunteers from the community."

Python 3.11 was released on 24 October 2022. The release notes put the result plainly: CPython 3.11 is "an average of 25% faster than CPython 3.10", as measured with the pyperformance benchmark suite, compiled with GCC on Ubuntu Linux. They are also careful about the spread: "Depending on your workload, the overall speedup could be 10-60%."

The myth it ended

One change matters to how people write code. Many developers avoided try blocks in hot loops "for speed". In 3.11, "'Zero-cost' exceptions are implemented, eliminating the cost of try statements when no exception is raised."

So in 3.11 and later, write the clear version:

import sys

print(sys.version_info >= (3, 11))   # this site runs Python 3.13

def parse(s):
    try:
        return int(s)
    except ValueError:
        return None

print([parse(x) for x in ["4", "x", "15"]])

True
[4, None, 15]

The try costs nothing on the rows that parse; you only pay when a ValueError is actually raised.

The part any team can copy

Before spending a week on optimisation, check which Python you run. An interpreter upgrade is usually the cheapest speed-up available: no code changes, and the gain applies to every line. Then measure your own workload, because "10-60%" is a range, not a promise. The performance lesson shows how to time code properly.