Tools · Text · Wörter und Zeichen zählen · Anleitungen · How to Count Words, Characters, and Reading Time (Live, No Upload)
How to Count Words, Characters, and Reading Time (Live, No Upload)
Paste or type into the word counter and seven numbers update as you go: words, characters, characters without spaces, sentences, paragraphs, reading time and speaking time. There is no Run button and nothing to submit - the count moves with every keystroke, about 150 ms after you stop typing.
Words, and why a single word without spaces still counts as one
The word count splits the text on whitespace after trimming the ends, so a pasted URL or a base64 blob with no spaces in it counts as one word, the same rule every counter of this kind uses. The exception is a script written without spaces - Chinese, Japanese, Thai - where a whole sentence would otherwise count as one word, so those runs are cut into words by the browser's own word segmenter. An empty box, or one with only spaces or line breaks in it, counts as zero words rather than one - a small guard, but the difference between a real zero and a false positive on a box you have not typed into yet.
Two character counts, because two different limits ask for them
"Characters" counts every character including spaces; "Characters (no spaces)" strips all whitespace first and counts what is left. Both counts are code-point aware - an emoji or an accented letter counts as one character, not as the two or more bytes it might take to store - which matters because some limits (a meta description, an SMS) are specified as character counts, not word counts, and the two numbers genuinely diverge on real text.
Sentences and paragraphs, counted by boundary rules
A sentence boundary is a run of . ! ? followed by a space or the end of the text, so a decimal like 3.5, node.js or example.com is not cut in two. The Arabic question mark ؟ and the Chinese or Japanese full stop 。 need no space after them - included because that is how those languages write. Unfinished text with no closing punctuation at all still counts as one sentence if it has any real content, so a paragraph you are still writing does not read as "0 sentences" while you are mid-thought. A paragraph is any block of text separated from the next by one or more blank lines; when the text has no blank line at all, every non-empty line counts as a paragraph, because that is what a paste from a word processor looks like. A trailing newline at the end of your text does not create a phantom empty paragraph.
A worked example, so the rules are checkable
Paste this exact two-paragraph fixture in and the counter reads 17 words, 91 characters, 74 characters without spaces, 4 sentences and 2 paragraphs:
This is a test. Is it working? Yes!
Here is a second paragraph with no closing punctuation
The sentence count is the clearest one to check by hand: three sentences end in visible punctuation ("This is a test.", "Is it working?", "Yes!"), and the second paragraph's unterminated fragment counts as a fourth sentence because it has real content after the last punctuation mark, which is the exact rule described above rather than a rounding quirk.
Reading time and speaking time, and where the numbers come from
Reading time is your word count divided by 200 words per minute, the commonly cited average silent-reading speed; speaking time uses 130 words per minute, the average spoken rate. Both round up to the next whole minute once the exact value passes 1 - anything under a full minute shows as "< 1 min" rather than "0 min", so a short paragraph does not read as having no reading time at all.
The one limit worth knowing
The input is capped at 200,000 characters, which is already a short novel's worth of text - well past any essay, article or script this tool is meant for. Past that point nothing is truncated and nothing you typed is lost; a one-time notice just lets you know the tool works best under that size, and the count keeps updating on the full text regardless.
Nothing you paste here is uploaded, not even briefly - the whole count happens in your browser, and it works the same way for English, Arabic, French, German and Spanish text.