How to Remove Comments from Python Code (3 Ways)

Need to strip the comments out of a Python file before sharing it, counting lines or cleaning up AI-generated code? Here are three ways that work — and one popular shortcut that quietly breaks your code.

Why not just use a regex?

The first thing most people try is deleting everything after a #:

import re
cleaned = re.sub(r'#.*', '', source)

It looks fine until your code contains a # inside a string. A URL like "https://example.com/#top" gets cut in half, and your file no longer runs:

# before
URL = "https://example.com/#top"  # base URL

# after the regex
URL = "https://example.com/

Python's own tokenizer knows the difference between a comment and a string, so use that instead.

Method 1: The tokenize module (standard library)

The tokenize module splits source code into the same tokens Python itself sees. Drop the COMMENT tokens and put the rest back together:

import io
import tokenize


def strip_comments(source):
    tokens = tokenize.generate_tokens(io.StringIO(source).readline)
    kept = [tok for tok in tokens if tok.type != tokenize.COMMENT]
    return tokenize.untokenize(kept)


with open("app.py") as f:
    print(strip_comments(f.read()))

Strings are untouched, so "https://example.com/#top" survives. Two things to know:

• Whitespace is left behind. untokenize keeps tokens in their original columns, so a trailing comment leaves trailing spaces, and a comment-only line becomes a blank line. Clean up with "\n".join(line.rstrip() for line in result.splitlines()).
• Docstrings are not comments. To Python they're string expressions, so tokenize keeps them. See method 3.

Method 2: A free online comment remover

If you just want clean code without writing a script, paste it into the Code Comment Remover. It reads your code the way a parser does, so:

• # inside strings and f-strings is never touched
• comment-only lines are removed completely — no leftover blank lines or trailing spaces
• tool directives like # noqa, # type: ignore and the shebang line are kept by default
• it also works for JavaScript, TypeScript, Java and C/C++

Method 3: Removing docstrings safely

Deleting docstrings by hand has a catch: if a function's only statement is its docstring, removing it leaves an empty body and a SyntaxError.

def not_implemented_yet():
    """TODO: write this."""

The safe approach is to find docstrings with the ast module (the first statement of a module, class or function that is a string constant) and replace a body that becomes empty with pass. The Comment Remover does exactly this when you tick Remove Python docstrings.

Should you remove comments at all?

For sharing a snippet, counting real lines of code or shipping a minimal script, yes. For code you'll keep maintaining, keep comments that explain why — business rules, workarounds, links to issues — and only remove the ones that repeat what the code already says. That's common in AI-generated code, and the Code Humanizer removes just those.

Summary

• Don't use re.sub(r'#.*', '', ...) — it breaks strings that contain #.
• Use tokenize for a scriptable, standard-library solution.
• Use the free Comment Remover for a one-click result with clean whitespace.
• Remove docstrings with ast, adding pass where a body would become empty.