Loading page
About this tool
Detect, validate, repair, preview, compare, copy, and download real browser-based file encoding conversions.
The File Encoding Conversion Tool reads a text file's raw bytes, detects its character encoding from the byte-order mark or byte patterns, and re-encodes it to UTF-8, UTF-8 with BOM, UTF-16 LE or BE, ASCII, Windows-1252 or ISO-8859-1. It decodes in strict mode first, so a file that is not actually valid in the chosen source encoding is reported rather than silently filled with replacement characters, and it can repair classic mojibake — text where UTF-8 bytes were read as Latin-1, producing sequences like é and ’. It is for developers, data engineers and anyone who has opened a CSV or log and found the accents turned into gibberish.
Open File Encoding Conversion Tool on AltFTool — it loads instantly in your browser.
Drop text, CSV, JSON, XML, HTML, CSS, JS, logs, or code files onto the Upload Panel; the Encoding Detection Panel then reports the Detected encoding, a Confidence percentage and any BOM found.
In the Encoding Selector pick a Source Encoding (Auto Detect, UTF-8, UTF-16 LE, Shift_JIS, GB18030, Big5, EUC-KR, Windows ANSI and more) and a Target Encoding (UTF-8, UTF-8 with BOM, UTF-16 LE, UTF-16 BE, ASCII, Windows ANSI or ISO-8859-1), ticking "Repair garbled UTF-8 text decoded as ANSI or ISO" if the preview shows mojibake.
Check the Validation Panel for errors, then use Copy Converted Text, Download Active File — saved as the original base name plus the target encoding, such as report_utf8.csv — or Batch Download All.
The detector checks for a BOM first, then tests ASCII, UTF-8 validity, UTF-16 patterns and single-byte Latin heuristics, and tells you how confident it is rather than just guessing silently.
Decoding runs in fatal mode, so an invalid source encoding raises an error, and any replacement characters in the result are counted and reported before you download.
If the text contains characters the chosen target cannot represent — anything non-Latin heading to ASCII or Windows-1252, for example — the conversion is blocked or warned about instead of quietly dropping them.
Turn on the mojibake repair: it takes each character's code point back to a single byte and re-decodes the sequence as UTF-8, which undoes the classic case of UTF-8 bytes having been read as Latin-1. It only applies when every character is below U+0100, so genuinely multilingual text is left alone.
It reads UTF-8, UTF-8 with BOM, UTF-16 LE and BE, GB18030, Big5, Shift_JIS, EUC-KR, ASCII, Windows-1252 and ISO-8859-1, or auto-detects. It writes UTF-8, UTF-8 with BOM, UTF-16 LE, UTF-16 BE, ASCII, Windows-1252 and ISO-8859-1.
Without, in almost every case. The three-byte BOM (EF BB BF) is unnecessary for UTF-8 and breaks shell scripts, JSON parsers and CSV importers that expect the first character to be data; add it only when a specific Windows application or Excel workflow requires it to recognise the file as Unicode.
Previews are capped at 9,000 characters for responsiveness — the conversion and the download still cover the whole file. The byte and character counts shown beside the preview are calculated over the entire input and output, so use those to confirm nothing was lost.