Text Encoding Inspector & Converter

How to inspect and convert a text file

1. Choose a text file. Leave source on Auto for BOM-marked Unicode or valid UTF-8, or explicitly select the encoding used by your source.
2. Choose UTF-8, UTF-16 little/big-endian, Windows-1252, exact Latin-1 or ASCII output. Select a BOM for Unicode output if required by your receiving application.
3. Preserve line endings or normalize CRLF, LF and CR to LF or CRLF. Choose Inspect & convert.
4. Review the source evidence, control-character counts and literal preview. Download the converted text or metadata report.

Detection evidence, not an encoding guess

Auto recognizes UTF-8 and UTF-16 BOM signatures. UTF-32 BOM signatures are recognized and rejected because UTF-32 is outside this tool's supported formats. Without a BOM, Auto accepts only strictly valid UTF-8. If validation fails, select the known source encoding explicitly.

ASCII-range bytes are valid UTF-8 and compatible with many other encodings. Some legacy-encoded files also happen to be valid UTF-8. Validity does not establish the author's intended encoding or prove that binary bytes are text. NUL and control-character counts help you review the result; they are not a binary-file detector.

Strict conversion and supported encodings

Malformed UTF-8, odd/truncated UTF-16, unpaired surrogates and non-ASCII bytes in ASCII mode stop decoding. A BOM that conflicts with an explicit source selection also stops conversion. Invalid data is not replaced silently. If a Unicode character is not representable in Windows-1252, Latin-1 or ASCII output, no converted download is created.

Latin-1 here maps bytes 0–255 directly to U+0000–U+00FF. It is distinct from Windows-1252, whose 0x80–0x9F range includes the euro sign and smart punctuation. The browser's Windows-1252 mapping is used for that encoding; Latin-1 is handled directly so browser label aliases do not change its meaning.

BOM and newline behavior

One matching input BOM is consumed as a signature. Additional U+FEFF characters remain content. Output BOM selection controls an additional leading signature for UTF-8/UTF-16 only. UTF-16 byte order is always explicit.

Preserve keeps CRLF, lone LF and lone CR exactly in the converted bytes. Normalization changes only those line-ending sequences; it does not modify Unicode line separators, whitespace, normalization forms or character content otherwise. The preview may not visibly distinguish every control character, so use the counts or inspect the downloaded bytes.

Limits and privacy

Input files are limited to 16 MiB and converted output to 64 MiB. Processing runs in a worker with a 15-second deadline. The preview is limited to 65,536 UTF-16 code units and never splits a surrogate pair. Downloads include the full output. Changes to settings or files invalidate the previous output; Cancel stops processing and Clear releases the selected file and download URLs.

No source bytes or converted text are uploaded, saved in browser storage, executed or rendered as HTML. The original file is never overwritten. Metadata reports contain the original filename and conversion statistics, which may still be private. This tool handles plain-text encoding only; it does not repackage DOCX, PDF, compressed files or other binary containers.

Encoding behavior is based on the WHATWG Encoding Standard and the browser's text decoding API.

Frequently asked questions

Continue your workflow