Convert a text file’s encoding

How it works

1. Drop the CSV or text files. Several at once are fine, even in different encodings.
2. Choose what to convert to. UTF-8 with BOM is what Excel opens without garbling; pick another if a system asks for it.
3. Leave the current encoding on Detect automatically unless you know better. The page reads the bytes and says what it found.
4. Press Convert and download. Each file keeps its name, and the page tells you if any character had no place in the new encoding.

Frequently asked questions

Why does my CSV look garbled in Excel?

Excel on Windows assumes a CSV is in the old encoding of your system (Windows-1252 in Western Europe, Shift_JIS in Japan) unless the file starts with a byte-order mark saying it is UTF-8. A CSV exported from Google Sheets, a web app or a Mac is usually UTF-8 without that mark, so accents and Japanese come out as é or 縺. Converting it to UTF-8 with BOM fixes it without losing a single character.

How is the current encoding detected?

A byte-order mark settles it at once. Without one, the page checks whether the bytes are valid UTF-8, which is almost never true by accident, and otherwise decodes them as each older encoding and keeps the one that reads most like real text: hiragana for Japanese, common syllables for Korean, accented letters inside words for European languages. If it guesses wrong, choose the encoding yourself.

What happens to characters the new encoding does not have?

They become a question mark, and the page lists them next to the file so nothing disappears silently. Emoji and rare kanji such as 𠮷 are the usual ones when converting to Shift_JIS. UTF-8 and UTF-16 hold every character, so nothing is lost in that direction.

Are the files uploaded?

No. The conversion runs in your browser. Customer lists and bank exports, the files people most often need to convert, never leave your device.

What do the line-break options do?

Windows ends lines with CRLF and Mac and Linux with LF. Most programs accept either, but some upload forms and older systems insist on CRLF. Keep as they are changes nothing.