Characters, Code Points and Bytes in One View
Type text to see how Unicode builds it: grapheme clusters, code points, UTF-8 bytes, UTF-16 units, HTML, CSS and JavaScript escapes and the four normalization forms, plus five common letter styles. These outputs are plain characters, not installed fonts or native rich-text formatting. Several maps overlap other TextToolbox tools; a name is not proof of a unique alphabet.
Start with the job. To browse all 53 copy-and-paste styles, use the Font Generator. To see what characters a piece of text is made of, or to get an escape you can paste into code, use this page. Search narrows the view names and descriptions; category and favourites filters narrow that selection further.
From text to a copied result
Enter the complete source you want to inspect, including spaces and accents; Sample offers short drafts and coverage checks. Compare the complete labeled outputs: search and category filters intersect, and you can choose favourites for this browser. Open an inline Preview when you want a larger separate field, and Copy uses the same cached value you see. If Copy fails, select the output and copy manually, and check the actual destination before publishing.
What each view shows
Input is segmented into complete source grapheme clusters. A mapping replaces a supported whole cluster; combined letters, non-Latin clusters and emoji remain unchanged in map-based rows. This does not normalize accents or transliterate a language. A substitution can still change subsequent boundaries or rendering.
Underline and Strikethrough append one combining mark after each non-whitespace cluster, so one visible letter becomes two code points. Code Points lists each Unicode code point as U+ followed by hexadecimal digits. A character outside the Basic Multilingual Plane, such as U+1F602, is one code point but two UTF-16 units and four UTF-8 bytes.
Grapheme Clusters separates the text into the characters a reader sees, so a letter with an accent or an emoji sequence counts as one. NFC, NFD, NFKC and NFKD are the four Unicode normalization forms. NFC joins a letter and its accent into one code point where possible, and NFD splits them into two. The K forms also fold compatibility lookalikes such as fullwidth letters into ordinary ones. Normalizing can make two strings that look the same compare as equal.
Binary and Hex encode the complete input as UTF-8 bytes, including whitespace; they are serialized representations rather than visual fonts. UTF-16 Units lists the 16-bit units a JavaScript string is made of, and Decimal Code Points gives the same numbers as Code Points in base ten.
JavaScript Escapes writes each character as \uXXXX, or \u{...} above U+FFFF, with short escapes for line breaks and tabs. HTML Entities writes numeric hexadecimal character references. CSS Escapes writes six-digit escapes for use in a content value. Test an escape in your own code or markup before relying on it.
Exact examples, including what NFD changes
| Input and style | Output | What to check |
|---|---|---|
| X ego, Bold | 𝗫 𝗲𝗴𝗼 | Capital X remains capital |
| ego, Script | ℯℊℴ | Letterlike script exceptions, not italic substitutions |
| é, Code Points | U+00E9 | One code point for the precomposed letter |
| é, Binary | 11000011 10101001 | Two UTF-8 bytes |
| é, NFD | é | Looks the same as é but is two code points: e and U+0301 |
| é, JavaScript Escapes | \u00E9 | Paste into a JavaScript string |
| A, HTML Entities | A | Numeric hexadecimal reference |
| e plus U+0301, mathematical map | é | Whole combined cluster is retained |
Choosing a view for the job
Use this page when you need to see what characters a text is made of, or to turn text into an escape for code. Binary and Hex help inspect UTF-8 bytes; they do not create a visual font. For decorative styles beyond the five letter styles here, use the Font Generator. Keep important names and instructions in ordinary text.
| Task | Starting view | Limitation |
|---|---|---|
| Compare letter shapes | Bold or Script; narrow with category or search | Mapped characters are different identities; combined letters can stay ordinary |
| Inspect characters | Code Points or Decimal Code Points | One entry per code point; one visible character can contain several |
| Compare two strings that look the same | NFC or NFD, then Code Points | Normalization changes code points, not what a reader sees |
| Inspect byte encoding | Binary or Hex | UTF-8 bytes of the entire input, including spaces |
| Paste into code or markup | JavaScript Escapes, HTML Entities or CSS Escapes | Test the result in your own environment |
| Keep a shortlist | Favourite rows, then Favourites only | Choices belong to this browser origin; they do not sync across devices |
Why valid Unicode can still be refused, and what to check
Unicode validity does not establish account-name permission, display consistency, search equivalence or identical speech output. Lookalikes are different characters. Font fallback, line wrapping and field filters can alter appearance or reject a result. Copying a style does not carry arbitrary color or size.
Input is capped at 2,000 UTF-16 units and 500 source clusters. Unsupported C0/DEL controls and incomplete surrogate sequences are rejected; tabs and line endings are allowed. In this Unicode grid, results are capped at 50,000 units each and 500,000 total; UTF-8 export size is checked before constructing its token string. Errors keep input editable and clear results without silent shortening. Browser segmentation data can differ.
Test supported letters and combined/non-Latin text together, inspect stored output counts, and retain the source. Favourites are validated known style names in website-origin browser storage; blocked storage keeps choices only during this visit. For the full set of 53 copy-and-paste styles, see the Font Generator. For what the site collects and how, read the Privacy Policy.
Reference character identities in the Unicode Mathematical Alphanumeric Symbols chart and Letterlike Symbols chart. Chart glyphs do not guarantee identical font artwork.