Japanese punctuation differs from Latin punctuation in two important ways: the
sentence-ending full stop is the small circle 。 (U+3002, maru) rather than a
dot, and there are no spaces between words. A counter built for English would
miss the 。 entirely and have no reliable word boundaries. This tool is built
for Japanese.
How it works
The text is split on runs of sentence-ending marks: the ideographic full stop
。, the full-width question mark ？, the full-width exclamation ！, their
ASCII equivalents ! ? ., and ellipses (…, ‥). Consecutive terminators
collapse into a single break so ！？ or a doubled …… ends just one sentence.
Each segment that contains kanji, kana, or alphanumeric content counts as one sentence. Because Japanese has no spaces, the tool reports characters rather than words, and gives characters-per-sentence as a length signal.
Example
The passage:
日本語は美しいです。話しますか？is two sentences: a statement closed by 。 and a question closed by ？. The
enumeration comma 、 inside a list would not add to the count.
Notes
- The maru
。is the primary sentence ending, not the dot. - The enumeration comma
、is a clause separator and is ignored. - Character counts are used instead of word counts because Japanese is unspaced.
- Everything runs locally; your text never leaves the browser.