Unicode Inspector
See every character in a string: code point, UTF-8 and UTF-16 bytes, category, script — with invisible, look-alike and combining characters flagged.
A per-character table, code point / byte / grapheme counts, whether the text is NFC-normalised, a script breakdown, and warnings for zero-width, bidi, unusual-space and mixed-script characters.
Example: "pаypal" pasted from an email lists a Cyrillic "а" (U+0430) among Latin letters and a zero-width space before the "p" — two things that make a lookalike domain and a failed search.
Every character,
named by its number.
Where the categories and scripts come from, what counts as flagged, and why names are not shown.
What is listed
The text is walked code point by code point (so an emoji is one row even though it is two UTF-16 units). Each row shows the U+ number, the UTF-8 bytes, the UTF-16 units, the Unicode general category (Lu, Ll, Nd, Zs, Cf …) and the script, both taken from the browser's own Unicode property support through \p{…} regular-expression escapes — so the Unicode version is whatever your browser ships. Grapheme clusters are counted with Intl.Segmenter.
Flags and normalisation
A character is flagged when it is invisible or easily mistaken: zero-width and word joiners, bidi controls (including U+202E, which reverses displayed text), no-break and other unusual spaces, soft hyphen, BOM, control, format, private-use and unassigned code points, and combining marks (which change the previous character). The text is also compared with its NFC, NFD, NFKC and NFKD forms: a string that is not NFC may look identical to one that is yet compare unequal, which is the usual cause of "the same" search term failing.
Limits
Character names ("LATIN SMALL LETTER E WITH ACUTE") are not shown — the name table is about 1.5 MB and would slow the page for everyone; the code point is enough to look one up. Only a fixed list of 29 scripts plus Common/Inherited is recognised; anything else is "Other". Mixed-script warnings are hints, not verdicts: a French sentence with a Greek letter is fine. Nothing leaves the browser; the same four anonymous usage counts as the rest of the site apply.
SOURCES
Last reviewed 20 September 2026. How results are checked: How we verify.