Skip to main content
ZEKILO Dev

Full-width Half-width Converter

Paste text to turn full-width letters and digits into half-width and half-width katakana into full-width.

Stays in your browserUAX #11As of 2026.10.03

Input

Press Esc, then Tab to leave the editor.

Output

How to use Full-width Half-width Converter

  1. 1Paste text into the input panel, or press Sample.
  2. 2Choose a direction for letters & digits, symbols, katakana, and space. The result updates as you go.
  3. 3Press Copy to copy the result, or Download to save it as a .txt file.

Full-width Half-width Converter options

Letters & digits
Full-width → half-width (default), Half-width → full-width, or Keep. Covers 0–9, A–Z, and a–z and their full-width forms.
Symbols
Full-width → half-width (default), Half-width → full-width, or Keep. Covers the ASCII symbols such as ! ? ( ) and their full-width forms, plus ¢ £ ¬  ̄ ¦ ¥ ₩.
Katakana
Half-width → full-width (default), Full-width → half-width, or Keep. Going to full-width, a voiced or semi-voiced mark is combined with the letter before it (ガ becomes ガ). Going to half-width, it is split again. The punctuation 。 「 」 、 ・ and the long vowel mark ー are included.
Space
Full-width → half-width (default) turns the ideographic space U+3000 into the ordinary space U+0020. Half-width → full-width does the opposite. Tabs and line breaks are never changed.

What is converted

Unicode keeps the full-width forms of ASCII (U+FF01 to U+FF5E) and the half-width katakana (U+FF61 to U+FF9F) in the block Halfwidth and Fullwidth Forms. Whether a character is wide or narrow is described in UAX #11, East Asian Width. This tool converts those characters and the ideographic space U+3000, grouped into four categories: letters and digits, symbols, katakana, and the space.

Each full-width ASCII character is exactly U+FEE0 above its ASCII counterpart, so A (U+FF21) becomes A (U+0041). Katakana use a table, because the order of half-width katakana differs from the order of full-width katakana.

Why not NFKC

NFKC normalization also folds full-width and half-width forms, but it changes much more: circled numbers, parenthesized Hangul and ideographs, unit symbols, ligatures, superscripts, and fractions all turn into plain characters. A tool for fixing the width of pasted text should not rewrite ① as 1. That is why the four categories here are mapped one by one and everything else is left untouched.

Full-width Half-width Converter examples

  • Full-width letters and digits, a full-width space, half-width katakana (defaults)

    Status line: letters/digits 6 · katakana 2 · space 1. Each voiced mark is combined with its letter.

    Input

    ABC123 ガギ

    Output

    ABC123 ガギ
  • Compatibility characters stay unchanged

    No characters changed.

    Input

    ① ㈜ ㎏

    Output

    ① ㈜ ㎏
  • Half-width to full-width

    Letters & digits: Half-width → full-width. Space: Half-width → full-width.

    Input

    ABC 123

    Output

    ABC 123

Examples use the reference cases this tool is tested against.

Full-width Half-width Converter: frequently asked questions

Is my input sent to a server?

No. Your input and the result are processed only in this browser and are not stored. Reloading the page clears them.

Why are ① and ㈜ not converted?

This tool does not use Unicode NFKC normalization. NFKC would also rewrite ① as 1, ㈜ as (주), ㎏ as kg, and fi as fi, which is rarely what you want. Only the four categories in the options are converted, each with its own mapping table.

How are voiced marks on half-width katakana handled?

カ followed by ゙ becomes the single character ガ, and ハ followed by ゚ becomes パ. A mark that cannot combine with the letter before it becomes a standalone ゛ or ゜. In the status line, a combined pair counts as one changed character.

Does it change hiragana, kanji, or Hangul?

No. Hiragana, kanji, Hangul, and emoji are left as they are. Full-width katakana that have no half-width form (ヰ ヱ ヮ ヵ ヶ) also stay unchanged.

What does the status line show?

The number of characters changed in each category, for example letters/digits 6 · katakana 2 · space 1. If nothing was changed, it says so.

How large can the text be?

Up to 1 MB is converted as you type. Above that, press Run to convert it. The maximum is 20 MB.

How it works · Standards

  • Processed in: this browser (your device). Nothing is sent or stored.
  • Standards: UAX #11
  • Engine: ZEKILO Dev width mapping tables (in-house code)
  • Line breaks, including CRLF, are kept as they are.
  • The half-width symbol forms U+FFE8 to U+FFEE (such as →) and arrows or box shapes are not converted in either direction.
  • As of 2026.10.03
  • Changelog: 2026.10.03 First release
Open-source licenses →
Something wrong with the result? Report an issue(Your input is not attached)