📖 Furigana Generator (ふりがな 付け)

Furigana (ふりがな) are small kana printed above or beside kanji to show their reading. They are essential for children learning to read, for learners of Japanese as a foreign language, and for any text that must be accessible to readers who do not recognise every kanji. Newspapers, easy-Japanese guides, karaoke lyrics, and e-book readers all rely on furigana. Without it, a word like 今日 can be read kyō, konnichi, or ima depending on context, and only furigana removes the guesswork.

This furigana generator scans your text, looks up each known kanji word in a built-in dictionary of common vocabulary, and inserts the matching reading as kana. The output keeps your original kanji and adds the furigana above them, so you can paste it straight into a document, a subtitle file, or an HTML ruby annotation. It is designed as a quick reading aid, not a full morphological analyser - for the script-side needs of the result, combine it with the kana converter and the romaji converter.

Refer to the table below for examples of how furigana appears above common kanji words:

The tool holds a dictionary that maps common kanji words to their kana reading. It scans the input from left to right and at each position tries the longest match it can find - so it prefers the three-character word 日本人 over the two-character 日本 plus the single character 人. When a match is found, the reading is attached to that span of kanji; characters and kana with no entry pass through unchanged. This is a dictionary lookup, not a full morphological parse, so rare words, proper nouns, and homophones with context-dependent readings may be missed or read with their most common form.

The table below summarises how the longest-match strategy resolves overlapping words:

Because the reading of a kanji can change with context (本 as a book is ほん, but 日本 is にほん), the tool always chooses the reading stored for the longest matched word. For the rare characters it does not know, the kanji is left as-is so you can fill in the reading by hand. To look up an individual kanji reading on demand, use the kanji reading lookup, and to count the characters in the output, drop it into the Japanese character counter.

1. Paste Japanese text that contains kanji into the input box.
2. Choose an output format - inline kana (word[reading]) or HTML ruby.
3. Click Generate to add furigana above each known kanji word.
4. Copy the result into your document, subtitle, or web page.
5. For any kanji left without a reading, look it up individually.

Is this a complete morphological analyser?

No. It is a dictionary lookup of common words using a longest-match strategy. It covers everyday vocabulary well but will miss rare words, jargon, and many proper nouns, which should be read by hand.

Why are some kanji left without a reading?

The built-in dictionary focuses on common vocabulary. Rare characters, technical terms, and names that are not in the dictionary pass through unchanged so you can add the reading yourself.

How does it choose between multiple readings of the same kanji?

It uses the reading stored for the longest matched word, which is usually the most common one. Context-dependent readings (like 明日 = あした vs あす) default to the most frequent form.

What output format should I use?

Use HTML ruby if you are pasting into a web page or rich-text editor, and inline kana (word[reading]) if you are working with plain text, subtitles, or spreadsheets.

Does it handle okurigana like 食べる and 読む?

Yes. Words that already include okurigana are matched as a unit, so 食べる is read たべる as a whole. To check okurigana spelling rules specifically, use the okurigana helper.

Is my text sent to a server?

No. The entire lookup runs locally in your browser. Nothing is uploaded, so it is safe to paste private or draft text.

Furigana is most useful for learners and children, but too much of it slows fluent readers - so target it to your audience. In HTML you can render furigana with the <ruby> tag, which this tool produces for easy pasting. Remember that some common words have multiple readings (明日 can be あした, あす, or みょうにち); this tool picks the most frequent one, so proofread context-sensitive cases. Proper nouns (names of people, places, brands) are not covered because their readings are often unique - run personal names through the kanji reading lookup instead. For punctuation-heavy text, the kana already present is left untouched, so mixed input works fine.