Text encoding detector and decoder Text encoding detector and decoder

Paste up to 20000 characters. A longer meaningful fragment makes detection more reliable.

Text encoding detector and decoder

Paste text that became unreadable after copying it from an email, old web page, file, CSV export or database. The service tries common mojibake repair paths and shows the most likely readable result.

The tool is useful when UTF-8 bytes were displayed as Windows-1251, Windows-1252 or another single-byte encoding. Always review short snippets manually because several variants can look plausible.

Restore text that became unreadable because the wrong character encoding was used.
The service checks common encoding pairs: UTF-8, Windows-1251, Windows-1252, ISO-8859, KOI8-R and CP866.
The best candidate is selected heuristically, while alternatives help compare similar results.
Useful for emails, legacy HTML pages, CSV files, logs, database dumps and imported documents.
A neural network is not required for most encoding errors: deterministic conversion is faster and easier to verify.
If the string is short or contains several languages, use the result as a hint and verify it manually.