Hiragana is one of Japan’s two kana syllabaries, used for native vocabulary, grammar particles, verb endings, and pronunciation guides. Each character stands for a syllable, so converting between Hiragana and the Latin alphabet is a matter of mapping syllable to syllable. This free tool applies the standard Hepburn romanisation in both directions, instantly and with no upload.
How it works
Going from Hiragana to Romaji, the tool reads the longest matching unit first so that contracted sounds (yōon) like きゃ are read as a single syllable kya rather than kiya. The small っ (sokuon) is detected and converted by doubling the consonant that follows it, turning がっこう into gakkou. Long vowels and the syllabic ん are handled by direct mapping.
Going from Romaji to Hiragana, the tool greedily matches the longest valid syllable at each position — so shi, chi, tsu, kya, and nn are recognised before single letters. A doubled consonant such as the kk in gakkou is turned back into a small っ followed by か, and trailing n becomes ん.
The five vowel rows
Hiragana is organised into rows by the consonant and columns by the vowel (a, i, u, e, o). The five core vowels are:
| Hiragana | Romaji | Sound |
|---|---|---|
| あ | a | ”ah” |
| い | i | ”ee” |
| う | u | ”oo” (unrounded) |
| え | e | ”eh” |
| お | o | ”oh” |
Hepburn special cases
Hepburn romanisation writes some syllables based on English phonetics rather than strict transliteration, which makes pronunciation easier for English speakers but can be surprising if you expect a purely systematic mapping:
| Hiragana | Hepburn | Why it differs |
|---|---|---|
| し | shi | Not “si” — English “sh” better captures the sound |
| ち | chi | Not “ti” — the sound is closer to “ch” |
| つ | tsu | Not “tu” — the cluster is a real affricate |
| ふ | fu | Not “hu” — bilabial fricative, closer to “f” |
| じ | ji | Not “zi” |
When typing Romaji for conversion back to Hiragana, use these Hepburn spellings —
shi, chi, tsu, fu, ji — for a clean round-trip.
Contracted sounds (yōon)
A small ゃ, ゅ, or よ following a row-i character creates a contracted two-kana sound written as one Romaji syllable:
| Hiragana pair | Romaji |
|---|---|
| きゃ | kya |
| しゅ | shu |
| にょ | nyo |
| ちょ | cho |
| りゃ | rya |
Geminate consonants (sokuon)
The small っ before a consonant doubles it in Romaji to mark a short pause or geminate:
- がっこう →
gakkou(school) - きって →
kitte(stamp) - ざっし →
zasshi(magazine)
Characters the table does not recognise — kanji, punctuation, numbers, and spaces — pass through unchanged. Everything runs locally in your browser.