Programming
Paste text to get its counts. Characters are reported three ways, because a tweet, a database column and JavaScript itself each count them differently.
1311311311033opinion3.5 characters1 min1 min4×3×2×1×1×1×1×1×1×1×This tool runs entirely in your browser. Your input is never uploaded, stored or logged.
How many characters is this? has three correct answers, and the gap between them is where the bugs are. Code points are what a person means by a character: 🚀 is one. UTF-16 code units are what JavaScript’s own .length returns: 🚀 is two, which is why slicing a string at a fixed index can cut an emoji in half and produce a broken character. UTF-8 bytes are what a VARCHAR limit, a network payload and an HTTP header actually count: 🚀 is four, and a Japanese character is three. All three are reported rather than picking one and being quietly wrong for the other two.
Words are found by matching runs of letters and digits, with an internal apostrophe kept as part of the word, so "don’t" is one word and "hello , world" is two. This works outside the Latin alphabet — Cyrillic and Greek text is counted the same way — which a whitespace split does not reliably do. Sentences end at a full stop, exclamation mark or question mark followed by whitespace or the end of the text, and a run of terminators counts once, so "Really?!" is one sentence rather than two.
Reading time uses 238 words per minute and speaking time 130, both from Brysbaert’s 2019 meta-analysis of reading rate studies rather than the 200 that most estimators repeat without a source. Both are rounded up, because a piece that takes 1.2 minutes to read does not take one. The frequency list ignores words of two letters or fewer, which would otherwise be entirely "the", "of" and "a", and ties are broken alphabetically so the same text always produces the same list.
Every example below is run against this tool in our test suite, so what it says here is what the tool actually does.
Last updated