How this tool works
The result separates Unicode code points from visible grapheme clusters and JavaScript UTF-16 code units. These are different measures, especially for emoji and combining accents. It also counts lines after normalizing line endings, blank-line-separated paragraphs and reading minutes at a stated 200-word-per-minute assumption.
Words = nonempty whitespace-separated tokens; reading minutes = words ÷ 200A worked example
“Hello world” has two words, eleven code points and one line. An emoji with skin tone or joined family members may occupy one visible cluster while containing several code points.
Good to know
Whitespace tokenization is explicit and is not a universal word-segmentation model for scripts without spaces. Reading time is an assumption rather than an observed reading speed. Browser segmentation support can affect cluster counts.
A character count needs a counting convention
JavaScript stores text in UTF-16 code units. Unicode code points describe encoded characters more directly, but a visible emoji or accented cluster can contain several of them. This counter displays these measures separately instead of labeling all of them simply “characters”. Browser grapheme segmentation groups common visible clusters together.
Words here are nonempty tokens separated by whitespace. That makes the rule reproducible for ordinary space-separated text, but it is not a linguistic segmentation engine for every writing system. The text case converter can change casing before you compare counts, and the bio line breaker can expose the effect of blank-looking separators.
Reading time and paragraphs are explicit approximations
Estimated reading minutes divide the token count by two hundred. Real reading speed depends on subject, language and reader, so this figure is a planning aid. A technical passage with formulas can take much longer than a familiar paragraph with the same token count.
Line endings are normalized before counting lines. A final newline contributes another empty line, while paragraphs are separated by blank lines rather than every line break. This distinction helps when pasted text contains hard-wrapped lines. The input remains in your browser for calculation; the page does not need a remote language-analysis service.
A few questions, answered
Does one emoji always count as one code point?
No. A joined or modified emoji can contain multiple code points but form one visible cluster.
Why does a final newline add a line?
It starts an empty line after the preceding text under this counter’s stated convention.