Your text
Inconsistencies found
A proofreading tool that finds and unifies hyoki-yure — inconsistent spellings — in Japanese text. Hyoki-yure is when the same word is written in more than one way: okurigana variants like 行う / 行なう and 申し込み / 申込み / 申込, kanji-vs-kana like 子ども / 子供 / こども and 下さい / ください, or katakana long-vowel variants like サーバー / サーバ and コンピューター / コンピュータ. Paste your text and the tool checks it against a built-in rule dictionary, then lists each inconsistent group together with how many times each form appears. Every group has "Unify to …" buttons; clicking one rewrites the whole text to that form (or use "Unify all" to normalize everything to the default form at once). Unifying respects containment between forms (サーバ is part of サーバー), so it never produces broken results like サーバーー. It's ideal for a final consistency pass over reports, manuals, web articles, or theses. The dictionary runs entirely in your browser, and the text you enter is never sent to any server. Because it only flags known variants from the dictionary, it won't silently "fix" proper nouns or technical terms you spelled intentionally.
How to use
- Paste the Japanese text you want to check into the input box (long text with line breaks is fine).
- Below it, inconsistent spelling groups appear with how many times each form occurs, grouped by type (okurigana, kanji/kana, katakana).
- Press the "Unify to …" button for the form you want, and the whole text is rewritten to it. "Unify all" normalizes everything to the default form at once.
FAQ
Is my text sent to a server?
No. Everything — including the rule dictionary used for detection — runs inside your browser. The text you enter is never uploaded, stored, or transmitted; it is processed only on your device, so it's safe even for drafts or confidential text.
What kinds of inconsistencies can it detect?
It detects well-known variants in its built-in dictionary: okurigana (行う / 行なう, 申し込み / 申込み / 申込), kanji vs kana (子ども / 子供, 下さい / ください, いただく / 頂く), and katakana long-vowel or v-sound variants (サーバー / サーバ, コンピューター / コンピュータ, ウイルス / ウィルス). Words not in the dictionary are not flagged.
What happens when I press "Unify"?
It rewrites every form in that group to the single form you chose. For example, if both サーバ and サーバー appear, pressing "Unify to サーバー" changes them all to サーバー. It handles containment between forms correctly, so it never produces broken text like サーバーー. The first form in each group (the default) is just a common-usage hint, not an absolute correct answer.
Could it flag a spelling that is actually correct?
The tool only reports a group when two or more of its forms appear in the same text. If you use only one form, nothing is reported. That said, proper nouns, product names, or quotations may use a spelling on purpose, so review the context before unifying.
Does it check full-width/half-width or alphanumeric inconsistencies?
This tool focuses on Japanese word-level inconsistencies (okurigana, kanji/kana, katakana). Full-width/half-width normalization and alphanumeric formatting are out of scope — use a dedicated kana / width converter for those.