Mojibake Fixer & Charset Converter

Auto repair

Paste any garbled text — 娴嬭瘯, 测试, é‚£ä½ — and the fixer tries common misread combinations (UTF-8 read as GBK, GBK read as Latin-1, and more), scoring each candidate by how natural the result looks. It iterates up to three rounds, so even double-encoded mojibake usually comes back clean. The repair path shows exactly which conversions were applied.

If the input contains replacement characters (�), the bytes were already lost before decoding; the tool tells you the repair can only be partial.

Manual conversion

Switch to Convert mode to re-interpret text between UTF-8, GBK/GB2312, GB18030, Big5, Shift-JIS, EUC-JP, EUC-KR, Latin-1, Windows-1252, and UTF-16 — with one click to swap directions. Encoding and decoding both run locally: decoding uses the browser's built-in decoders, and encoding uses lookup tables derived from them at runtime.

Byte inspector

Each side shows a hex dump (UTF-8 bytes) of the text so you can see exactly which bytes produced the garbling — handy when debugging APIs, databases, or CSV imports.

Your mode and charset choices are saved in this browser. All processing happens locally.