Paste any Japanese text and add furigana — the small kana reading hints — to every kanji.
Copy the result as ruby HTML, bracket style, or a full kana reading.
Your text never leaves your device — analysis runs in your browser.
Converting…
For web pages: <ruby>漢字<rt>かんじ</rt></ruby>
For Word, Google Docs, and plain text: 漢字(かんじ)
How to use
Paste Japanese text into the box above — an article, a song lyric, homework, anything.
Click Add furigana. The first conversion downloads a Japanese dictionary, so it can take 5–15 seconds. After that, conversions are instant.
Check the result in Preview, then copy it from the Ruby HTML, Bracket style, or Reading tab.
Output formats
Preview — the text rendered with furigana above each kanji, exactly as it would appear on a web page.
Ruby HTML — <ruby> markup you can paste into any web page. All modern browsers display it with the reading above the kanji.
Bracket style — 漢字(かんじ) plain text. Use this for Word, Google Docs, Anki cards, or anywhere HTML doesn't render.
Reading — the entire text converted to kana, with a hiragana / katakana toggle. Handy for checking how a whole sentence is pronounced.
About this tool
Every kanji gets furigana. Readings are estimated from context by morphological analysis, so compound words and verbs with okurigana are handled correctly (only the kanji stem gets the reading).
Everything runs in your browser. After the one-time dictionary download, your text is analyzed entirely on your device and never uploaded.
Free, no sign-up, no limits on text length within reason.
FAQ
What is furigana?
Furigana are small kana characters written above (or beside) kanji to show their pronunciation. Japanese children’s books, textbooks, manga, and learner materials all use furigana so readers can pronounce kanji they haven’t learned yet. On the web, furigana is displayed with the HTML <ruby> tag.
How do I use furigana in Word or Google Docs?
Ruby HTML only renders on web pages, so pasting the HTML code into Word or Google Docs will show the raw tags instead of stacked readings. For documents, use the Bracket style output — 漢字(かんじ) — which works anywhere plain text does. Microsoft Word also has a built-in “Phonetic Guide” feature if you need true ruby layout inside a document.
Is my text uploaded to a server?
No. On the first conversion your browser downloads a Japanese dictionary from a CDN, and from then on all analysis runs locally on your device. The text you paste is never sent anywhere.
Why is a reading sometimes wrong?
Readings are estimated from context using morphological analysis (the kuromoji.js library). It is accurate for ordinary text, but personal names, place names, and words with irregular readings can be misread. The output is plain text and HTML, so you can copy it and fix any reading by hand.