Live counts as you type, plus readability, keyword density, platform limits and a writing check that highlights long sentences, passive voice and filler words. Everything stays in your browser.
Every word counter gives you a number. The useful question is why your number disagrees with someone else's. Microsoft Word, Google Docs, this tool and your lecturer's marking rubric can all look at the same essay and report different totals, because each one draws the line in a different place around hyphens, numbers, footnotes and captions. This page tells you exactly where the lines are drawn here, so when a limit is tight you know what you are working with.
Alongside the counts you get readability scores, keyword density, live platform character limits, and a writing check that highlights long sentences, passive voice, adverbs and filler words directly in your text. Everything runs in this browser tab — your draft is never uploaded, which matters if you are pasting in unpublished work, client copy or anything under NDA.
| Metric | Rule used here |
|---|---|
| Words | Runs of letters and digits. Apostrophes and internal hyphens stay joined, so don't and state-of-the-art are one word each. |
| Characters | Every character including spaces, punctuation and line breaks. Emoji count as one. |
| Sentences | Split on . ! ? and line breaks. Any run of text with at least one word counts, even without a full stop. |
| Paragraphs | Separated by a blank line, matching how most word processors export text. |
| Reading time | 238 words per minute — the mean for adults reading non-fiction silently. |
| Speaking time | 140 words per minute, a comfortable presentation pace. |
X and Bluesky, Threads, LinkedIn posts and headlines, Instagram captions, YouTube titles and descriptions, SMS segments, plus the two that matter for search: a 60-character title tag and a 155-character meta description. Each shows a bar that turns amber near the limit and red past it.
Because Word counts some things this tool does not, and vice versa. Word includes text inside footnotes, endnotes, headers, footers and text boxes in its main count, and it treats a number like 2026 and a lone symbol as words. It also counts an em-dash-joined construction differently depending on spacing. Differences of 1–3% between any two counters are completely normal. If a hard limit is close, count with whatever tool your marker or editor will use.
One, here and in Word. Hyphens inside a word keep it joined, which is why hyphenated compounds do not inflate your count. An em dash or a spaced hyphen does separate words — well - known is two, well-known is one. This is worth knowing when you are a few words over a limit.
One. The apostrophe is treated as part of the word, so don't, it's and O'Brien each count once. Expanding contractions therefore increases your word count — a genuinely useful trick when you are just under a minimum.
With spaces counts every keystroke including the space bar, tabs and line breaks. Without spaces counts only the visible characters. Which one you need depends who is asking: translation and copywriting rates are usually quoted per character with spaces, while some database fields and older publishing systems count without. Social platforms count with spaces.
Yes. X counts every character including spaces and line breaks against the 280 limit. Links are the exception — every URL is shortened to a fixed 23 characters regardless of its real length, so a very long link costs you the same as a short one. Images and quoted posts do not consume characters.
It is an estimate based on 238 words per minute, which is the average silent reading speed for adults on non-fiction, drawn from a well-known meta-analysis of reading research. Real speed varies enormously: technical or unfamiliar material can drop below 150 wpm, light fiction can exceed 300. Treat it as a ballpark for a blog post, not a promise.
Higher is easier. 60–70 is the usual target for general-audience writing and is roughly what a mainstream newspaper hits. Above 80 reads as simple and direct; below 30 is dense academic or legal prose. Note that the formula only looks at sentence length and syllables per word — it has no idea whether your writing is good. A page of short, punchy nonsense scores beautifully.
Flesch–Kincaid grade level maps your text onto US school years, so 8.0 means an average 13–14 year old should follow it. Most public-facing writing aims for grade 8–10; UK government guidance recommends around grade 9 for public information. Very short simple sentences can produce a score below zero, which just means "easier than the scale goes".
By counting vowel clusters and applying rules for silent endings, which is the standard approach. It is approximate by design — English spelling is not phonetic, so words like science, area and fire can be off by one. Across a paragraph these errors largely cancel out, which is why the readability scores stay useful even though individual words occasionally miscount.
It is how often a word or phrase appears as a percentage of total words. It matters far less than it did fifteen years ago — Google stopped rewarding repetition long ago and actively penalises stuffing. Use it as a sanity check: if your main term is at 5% you are almost certainly writing awkwardly, and if it appears twice in 2,000 words the page may genuinely be about something else. Somewhere around 0.5–2% usually reads naturally.
Because "the", "and" and "of" are always the most frequent words in English and tell you nothing. They are filtered by default so the list shows terms that actually characterise your text. Switch the stop-word filter off if you want the raw frequency, which is occasionally useful for spotting a tic — overusing "however" or "just", for instance.
The check flags a form of to be followed by a past participle: "was written", "is being reviewed", "has been decided". No, it is not always wrong. Passive voice is the right choice when the actor is unknown, irrelevant or deliberately de-emphasised — "the samples were incubated for 48 hours" is correct scientific style. Treat the highlight as a prompt to check, not an instruction to rewrite.
Words ending in -ly often signal a weak verb propping itself up: "ran quickly" is usually better as "sprinted". That is a stylistic convention popularised by Hemingway-style editors, not a grammatical rule. Plenty of -ly adverbs are essential — "legally", "annually", "genetically" carry real meaning. The tool skips obvious false positives like "only", "family" and "reply".
Over 20 words is flagged amber and over 30 red. These are conventions, not laws. Long sentences are harder to follow, especially on screens, but a well-built long sentence with clear structure beats a run of choppy short ones. The value of the highlight is spotting a paragraph where every sentence is long — that is where readers give up.
Intensifiers and hedges that usually add nothing: very, really, quite, just, actually, basically, literally, simply, obviously, definitely and similar. They are not errors and in speech they are natural, but in writing they typically dilute the sentence. "This is really very important" says less than "this matters".
Yes — set your target in the Word goal panel and it shows the remaining count, with a bar that turns red when you go over. Two cautions: check whether your institution counts the title, abstract, footnotes, tables, captions and reference list, because policies differ wildly, and confirm the final figure in the software your marker will use. A 10% tolerance either side of the limit is common but never assume it.
This tool counts every word you paste, including the bibliography. Most universities exclude the reference list but include in-text citations such as (Smith, 2024). Since practice varies by department and even by module, the reliable method is to paste only the body text you believe counts, then check the number against your handbook.
Each CJK character counts as one word, which is how Microsoft Word behaves and how most publishers count these languages. This is the right convention because CJK writing does not separate words with spaces, so counting space-delimited runs would report a whole paragraph as a single word. Latin text mixed into the same document is still counted normally.
As one character each, which matches how a person would count them. Technically many emoji occupy two units of storage, and some — flags, and people with skin-tone or profession modifiers — are built from several joined together. Some platforms bill those as multiple characters against a limit, so if you are right on the edge of 280 on X with several complex emoji, verify in the app itself.
A single SMS holds 160 characters. Go one over and the message splits into segments of 153 characters each, because space is taken by the header that reassembles them. Since you are usually billed per segment, a 161-character message costs double a 160-character one. Any non-GSM character — most emoji, curly quotes, many accents — switches the whole message to a 70-character-per-segment encoding.
Google truncates based on pixel width, not character count, so there is no exact limit — a title full of narrow letters fits more than one full of capitals. Around 60 characters and 155 characters are the practical thresholds where most titles and descriptions survive intact on desktop. Mobile truncates sooner. Front-load the important words and nothing is lost when it cuts.
No. All counting happens in this browser tab. Nothing is transmitted to a server, so there is no copy of your draft anywhere but your own machine — which matters for unpublished manuscripts, client work and anything confidential. Your text is saved to this browser's local storage so it survives a refresh; press Clear to remove it, and it never leaves the device either way.
Yes, once the page has loaded. There are no external libraries doing the counting — it is all plain JavaScript in the page — so you can disconnect and keep working. That is also the simplest way to satisfy yourself that nothing is being sent anywhere.
Not directly — you can drop in .txt and .md files, or paste from anything. For a .docx or .pdf, opening it and copying the text works fine and takes a moment. Pasting also has an advantage: you see exactly what you are counting, rather than trusting an extractor that may pull in headers, footers and page numbers you did not intend to include.
No imposed limit. A 20,000-word document analyses in well under a tenth of a second, so a full dissertation is comfortable. Very large pastes — hundreds of thousands of words — will eventually feel sluggish because the writing check re-scans on every keystroke; turn the check off with the toggle and counting stays instant.
Free, no account, no ads, no watermark and no limits. It is one of a set of browser tools on this site. Analytics are privacy-first and cookie-free, and record only that a page was viewed — never what you type.