Word, character and reading time counter
Paste your text and you immediately see words, characters with and without spaces, sentences, paragraphs, unique words, average sentence and word length, reading time and speaking time, manuscript pages and print pages. There is no "Calculate" button: the numbers update on every keystroke. This is a working tool for an author, not a toy. English-language publishing counts in words: an editor asks for a word count, a magazine buys by the word, an agent thinks in standard manuscript pages of 250 words, a publisher estimates the finished book in printed pages, and the reader measures text in minutes. The same piece of writing looks different in each of these systems, and converting between them in your head every time is wasted effort.
Enter some text and the explanation of the maths appears here.
Longest sentence
Enter some text and the longest sentence appears here.
Its length appears together with the sentence.
Sentence boundaries are approximate. The tool understands the full stop, the exclamation mark, the question mark and the ellipsis, and it tries not to split on abbreviations such as "etc." or "Dr.", but there is no fully reliable way to split text into sentences without understanding what it says.
Your text never leaves your browser: neither a draft nor a finished article is sent anywhere or stored anywhere.
What is counted here, and how
A word is a run of letters and digits. A hyphen or an apostrophe inside a word does not break it, so "part-time" and "don't" each count as one word, not two. This matters: many counters simply split on spaces and inflate the result on hyphenated words. The word is the primary unit here, because it is the one English-language authors, editors and publishers actually quote. Everything measured in pages is derived from the word count. Characters with spaces is the full length of the text, line breaks included. Characters without spaces is the same figure minus every whitespace character. For English the gap between the two is usually 15 to 18%, and it is exactly what starts arguments about length with a client. A paragraph is a block of lines separated by an empty line. Unique words are counted as word forms: "write" and "writes" are two different words, because the tool does no morphological analysis. Reading time is words divided by reading speed; speaking time is words divided by speaking speed. Both speeds are editable fields.
The manuscript page, the print page and why 250 words
A standard manuscript page is 250 words. The figure comes from the typewriter era: a page of pica type, double-spaced, 12 point, with one-inch margins, held about 25 lines of ten words. Agents, editors and competition rules still use it, because it is a unit of the author's work that does not move when somebody changes the font. That is why the words-per-page field is editable here rather than baked into the code — some houses count 275, some count 300. A print page is a different thing entirely: it is a page of the finished book. A trade paperback page holds roughly 300 words, so this tool divides by 300 to estimate how thick the book will be. It is only an estimate. Type size, leading, trim size, dialogue and white space all change the real page count, and a heavily dialogued novel takes far more pages than the same word count of dense non-fiction. Characters are the third system, and they are the one to be careful with. Converting characters to words depends on the language, the style and the subject, so it is always an approximation: an English word plus its space runs about 5.8 characters, a Ukrainian or Russian one closer to 6.5. Freelance marketplaces in Eastern Europe still price "per thousand characters" without spaces, while English-speaking clients pay per word. Before you name a price, look at both figures at once instead of converting by hand.
Why sentences are an approximation
The tool treats a full stop, an exclamation mark, a question mark and an ellipsis as a sentence boundary when a space or the end of the text follows. Then it checks the context: if an abbreviation stands before the stop ("etc.", "Dr.", "vs.", "fig.") or a single capital letter does (the initial in "J. Smith"), or if the next word starts in lower case, no boundary is placed. Cyrillic pairs such as "т. д." and "н. е." are handled separately: inside them the sentence is not broken, but before a capital letter it is, because that is where the thought really ends. It is still an approximation. Splitting text into sentences exactly is impossible without understanding the meaning, and every counter gets it wrong sometimes. So read the average sentence length as a guide, not as a measurement. The longest sentence is shown in full — the most useful output in the whole block. A long sentence is not a mistake, but it is the first place worth looking at when you edit.
What the counter does not do
It does not judge quality. A word count says nothing about whether the text is worth reading: an empty 3000-word piece stays empty, and a strong 400-word note stays strong. It does no morphological analysis, so it does not reduce word forms to a base form and does not count parts of speech. Keyword density and filler-word analysis have separate tools. It does not store your text. Nothing is sent to a server, nothing stays in the browser after you reload the page — the calculation lives in the tab's memory and goes away with it. And it does not reformat anything: markup, tags and extra spaces are counted exactly as you pasted them. If you copied an article as HTML, the tags land in the character count too.