Invisible Character Revealer

What you can do

• Find zero-width spaces and other hidden characters in a text.
• Detect zero-width watermarks in AI-generated or copied text.
• See non-breaking spaces and every other Unicode space variant.
• Spot bidirectional control characters that reorder text.
• Check whether a file uses LF or CRLF line endings, and convert between them.
• Remove all invisible characters with one click.
• Normalize special spaces to regular spaces.
• Find tabs and trailing whitespace before a code review flags them.
• Inspect any hidden character's name, code point and UTF-8 bytes.

How to use the Invisible Character Revealer

1. 1Paste your text into the input box, or press Try an example.
2. 2Read the findings: counts per kind, the line-ending style, tabs and trailing whitespace.
3. 3Look at the revealed text. Every hidden character shows up as a colored chip, and clicking one shows its code point, UTF-8 bytes and position.
4. 4Pick the cleanup fixes you want, or press Remove everything invisible.
5. 5Copy or download the cleaned text, or send it on to another tool.

What it detects

The catalog covers the zero-width family (zero width space, zero width joiner and non-joiner, word joiner, the byte order mark, variation selectors, invisible math operators and Unicode tag characters), all twelve bidirectional controls from the marks to the isolates, the full space family from the no-break space to the ideographic space, the soft hyphen, Hangul fillers, the braille blank and every C0 and C1 control character. Tabs, trailing whitespace and the LF versus CRLF question are reported separately, since they are ordinary characters in the wrong place rather than hidden ones.

Zero-width characters are also what most text watermarks and homograph tricks are made of, so a text that looks clean but carries a pattern of ZWSP and ZWNJ chips is worth a closer look.

How the cleanup works

Each fix targets one kind of character. Zero-width characters, bidi controls and control characters are removed outright, special spaces become regular spaces, and the Unicode line and paragraph separators become normal line feeds so the text keeps its structure. Trailing whitespace and line-ending conversion are separate switches, so you can fix a mixed-endings file without touching anything else. The counters above the result confirm when nothing invisible is left.

If the text looks encoded rather than hidden, the Magic Decoder can work out what was applied to it, and Find and Replace handles substitutions the cleanup options do not cover.