Text Encoding Detector

When the result is legacy or unknown

Text that is not valid UTF-8 is most often in an older single-byte encoding. For Western European text that usually means Windows-1252 or ISO-8859-1. Convert it once and keep UTF-8 from then on: iconv -f WINDOWS-1252 -t UTF-8 in.txt > out.txt on Linux or macOS, or open it in an editor that lets you choose the encoding and save as UTF-8. Only the first 1 MB is inspected, so a large file whose first megabyte is plain ASCII can still contain non-ASCII text further on.

Byte-order marks and UTF-16

A UTF-8 byte-order mark (EF BB BF) helps Excel open a CSV with accents correctly, but it breaks other things: a shell script's #! line, PHP output sent before headers, and the first column name of a CSV read by code that does not strip it. UTF-16 files usually come from Windows, for example from Windows PowerShell 5.1's > redirect. Many Unix tools read them as text full of null bytes.

How to use it

1. Choose a text-like file up to 50 MB.
2. Inspect the first 1 MB locally.
3. Review the encoding and confidence label.

Privacy & limitations

The file stays local. Encoding detection without a BOM is heuristic and cannot always distinguish legacy character sets.