pypi / charset-normalizer

charset-normalizer

What changed in charset-normalizer 3, from its official release notes, upgrade guides and advisories. Sources checked 2026-10-17.

5 breaking4 performance2 behavior and defaults4 tooling14 fixes and other

See which of these apply to your code, ranked, in your agent:

npx @unstale/cli init

charset-normalizer 329 changes

Large content with high-entropy noise detector output is no longer decoded, affecting inputs larger than 1M bytes.
3.5.1release notes
Files containing both a charset declaration and BOM/SIG now verify the BOM/SIG charset first.
3.5.0release notes
Removed in favor of feed_info, requiring callers to use the replacement API.
3.4.6release notes
Removed in favor of feed_info, requiring callers to use the replacement API.
3.4.6release notes
Replaces eligible and feed as the preferred API for the flattened logic.
3.4.6release notes

CharsetMatches no longer sort on each match insertion, reducing repeated sorting work.
3.5.0release notes
Cascade codec imports were changed to reduce wall import time.
3.4.8release notes
Measurements no longer use a global cache, reducing RSS memory usage.
3.4.8release notes
Noise and coherence measurements received micro-optimizations.
3.4.8release notes

The algorithm no longer uses regex search by default.
3.4.8release notes
Updated using Unicode blocks v17, changing the Unicode range data used by charset detection.
3.4.6release notes

The Cython upper bound is now <3.4 for native builds; abi3 builds remain capped below 3.3 for Python 3.7 Limited API compatibility.
3.5.2release notes
abi3 wheels are now distributed, reducing the need to rush releases for new Python interpreters.
3.5.0release notes
Optional mypyc builds were replaced with Cython extensions while retaining the pure Python fallback.
3.5.0release notes
Pre-built freethreaded wheels are now shipped in PyPI for freethreaded Python environments.
3.4.6release notes

Supported encodings without aliases now resolve correctly and are no longer ignored in charset declarations.
3.5.2release notes
Valid UTF-8 Chinese JSON is no longer incorrectly detected as PTCP154 because of excessive noise penalties for uncommon CJK characters.
3.5.2release notes
Comparisons with non-alias encoding strings were fixed.
3.5.0release notes
Empty payloads now return 0.0 instead of crashing.
3.5.0release notes
Detection of iso2022* cases was corrected by fixing multibyte chunking logic.
3.5.0release notes
Explicit support for Python 3.15 was added.
3.5.0release notes
Fixed a regression in the fallback path that could lead to a decode error.
3.4.9release notes
Inverse capitalization is now recognized by the noise detector.
3.4.8release notes
The redundant UTF7 BOM marker was removed.
3.4.8release notes
Correctly removes the SIG remnant from utf-7 decoded strings.
3.4.7release notes
Fixed an edge case where noise differences between candidates could be almost insignificant.
3.4.6release notes
Fixed writing to the wrong path when normalizing multiple files.
3.4.6release notes
Passing a bytearray to from_bytes no longer raises TypeError.
3.4.5release notes
Logger levels are restored correctly when processing empty byte sequences.
3.4.5release notes