Fullwidth / halfwidth converter
Convert between fullwidth and halfwidth characters: letters, digits, punctuation, the ideographic space and Japanese katakana (カ ↔ ｶ, ガ ↔ ｶﾞ), with spaces and punctuation as separate options.
Runs in your browserEvery computation happens in your browser — your data never leaves this device.
What this tool does
- Normalise fullwidth letters and digits pasted out of PDFs, spreadsheets or legacy systems: Ｈｅｌｌｏ，世界！ １２３ → Hello,世界! 123.
- Prepare halfwidth text for CJK typography: converting punctuation and digits to fullwidth makes them line up with Han characters.
- Handle Japanese katakana both ways — カタカナ ↔ ｶﾀｶﾅ — with voiced kana such as ガ and パ decomposed into base plus dakuten per the Unicode rules.
- Convert only part of the text: turn off the space or punctuation switch and those characters stay exactly as they were.
Example
Input
Ｈｅｌｌｏ，世界！ カタカナ １２３
Output
Hello,世界! ｶﾀｶﾅ 123
Fullwidth ASCII, the ideographic space (U+3000) and katakana are converted; ヵ ヶ ヮ ヰ ヱ ヽ ヾ have no halfwidth form and are kept unchanged.
Frequently asked questions
How does the fullwidth/halfwidth conversion actually work?
The fullwidth ASCII block U+FF01–U+FF5E and halfwidth U+0021–U+007E differ by a fixed offset of 0xFEE0, so they map by plain addition. The ideographic space U+3000 maps to a normal space, and katakana plus punctuation go through the Unicode compatibility mapping table.
Why did ，become a plain comma instead of 、?
Because this is a width conversion, not a punctuation-style conversion. The fullwidth comma U+FF0C sits inside the fullwidth ASCII block, so subtracting the offset gives the halfwidth comma. Use a dedicated Chinese punctuation tool if you want semantic replacements.
What happens if I turn off "convert punctuation"?
Letters and digits still convert, but punctuation is skipped: fullwidth ，。！？ stay fullwidth, halfwidth ,.!? stay halfwidth, and the Japanese ｡｢｣､･ ↔ 。「」、・ pairs are left alone. Handy when you only want to strip fullwidth alphanumerics without touching punctuation.
Why does a katakana character become two characters?
Voiced kana such as ガ, パ and ヴ decompose into a base kana plus a dakuten mark, and halfwidth has no single-character form, so ガ turns into ｶﾞ. Converting back recombines ｶﾞ into ガ, so the length shrinks again.
Does the conversion change meaning, or upload my text?
It only remaps character widths and never changes meaning — the result is still valid UTF-8. Everything runs in JavaScript inside your browser, the page sends no requests, and your text is never uploaded.
Keywords：全角半角fullwidthhalfwidth转换convert假名katakana日文标点