Chinese Mojibake Fixer （乱码修复）

Paste garbled Chinese text such as 浣犲ソ, 缃戠粶 or 锟斤拷 and this tool works out which GBK / Big5 / UTF-8 misread chain produced it, then repairs it with one click. The best candidate is shown with an honest three-state label: Fixed, Approximate or Irreversible — nothing is silently guessed.
All processing happens locally in your browser using a bundled charset table — your text is never uploaded to any server.
Garbled Text (乱码)
Repaired Text
Ctrl+Enter fixes.
Paste some garbled Chinese text and click Fix Now. All processing stays in your browser.
How it works: the tool encodes your text back with each charset, decodes the bytes with the other charsets, and scores every candidate by the number of valid common characters and replacement symbols. Typical chains it detects: UTF-8 bytes read as GBK (你好 → 浣犲ソ), UTF-8 read as Big5, and GBK read as UTF-8 (which leaves � or 锟斤拷). Fixed means a confident, clean result was found; Approximate means the character set round-trips but the result still contains rare or odd characters — proofread it; Irreversible means the original characters were damaged (replaced by �/锟斤拷) and cannot be recovered. For raw file bytes use the File Encoding Converter or the Chinese Charset Converter.
What is mojibake?

Mojibake (乱码) is text that was decoded with the wrong character set: the bytes themselves are intact, only the label reading them was wrong. Because nothing was destroyed, encoding the text back and re-decoding it with the right set can undo it — that round trip is exactly how this page repairs 浣犲ソ back to 你好.

Damage becomes permanent only when the misread text was saved and the wrong characters replaced the originals — the 锟斤拷 pattern is what the UTF-8 replacement bytes for already-lost characters look like when read as GBK.

†at the start of a garbled word is the signature of UTF-8 bytes read as Latin-1/Windows-1252; 浣犲ソ signals UTF-8 bytes read as GBK; a clean result here means the round trip was verified, not guessed.