Unicode Code Point Inspector

Inspect Unicode code points, zero-based UTF-16 offsets and code-unit lengths without confusing emoji with single code units.

Text processing stays in this browser. Input is limited to 20,000 UTF-16 code units.

Each row is one Unicode code point. Positions start at 1; UTF-16 offsets start at 0. JSON display quotes whitespace and controls.

See what a string actually contains

A character that looks like one symbol can contain several Unicode code points. This inspector separates those points and reports their hexadecimal values, their UTF-16 storage lengths and their starting offsets. It helps investigate differences in pasted labels, emoji or accented text without guessing from the font. It does not identify official Unicode character names, detect malicious text or certify that a string is safe.

Inspect an example

  1. Choose Load example (replaces input and settings) or paste into Input text.
  2. Select Process text. There are no inspection settings to configure.
  3. Read the report preview. Copy full result and Download result include every row, even when the preview stops after 200 rows.

The example contains A😀é: the last visible letter is an e followed by a combining acute accent. It produces four rows:

Position UTF-16 offset Units Code point
1 0 1 U+0041
2 1 2 U+1F600
3 3 1 U+0065
4 4 1 U+0301

The emoji occupies two UTF-16 code units but one code point. The final accent has its own row although it usually appears attached to e. Offsets are zero-based positions in the received JavaScript string, not byte offsets in a UTF-8 file. Position is a one-based report row number.

Read whitespace and exported rows

The JSON display column wraps each point in quotes. Tabs and line feeds use escaped forms so they cannot create extra TSV columns or records. Other invisible characters may still look blank in that column; the U+ value is the reliable identifier. Downloaded reports use tab-separated columns with a header. Spreadsheet software may interpret imported cells, so use a text editor when exact report strings matter.

Input is capped at 20,000 UTF-16 code units. Whitespace-only text is accepted; completely empty text is not. Unpaired surrogate code units are rejected rather than silently turned into replacement characters during UTF-8 export. A pasted textarea may normalize CRLF or CR line endings to LF, so offsets refer to the text received here, not the original file bytes.

When a code-point report is not enough

A sequence of points is not a measure of its visual width. Font shaping and writing systems affect what appears on screen. To group combining marks or joined emoji, use the Grapheme Cluster Viewer. To compare composed and decomposed accents, use the Unicode Normalization Tool. The distinction between UTF-16 units and code points is documented by MDN's codePointAt reference.