Word counter
Words, characters, sentences, paragraphs and reading time, updating as you type. Hyphenated compounds, contractions and URLs each count as one word, abbreviations do not end sentences, and Chinese and Japanese are counted by character — the cases that make other counters disagree with each other.
Nothing is uploaded, which matters when the text is a manuscript, a dissertation or a release under embargo.
Nothing is uploaded. Counted in this tab. No request carries what you paste.
Why two counters disagree
Word counting looks like splitting on spaces, and that is exactly why no two tools agree. The disagreements are all in the same handful of places.
| Text | Naive split | Here |
|---|---|---|
well-known | 1 or 2 | 1 |
mother-in-law | 1 or 3 | 1 |
don't | 1 or 2 | 1 |
https://example.com/a/b/c | 1 or 6 | 1 |
3.14159 | 1 or 2 | 1 |
你好世界 | 1 | 4 |
The scripts without spaces
Chinese and Japanese are written without spaces between words. A counter that splits on whitespace therefore reports one word for an entire page of Chinese, which is not a rounding error — it is wrong by a factor of several hundred.
Those languages are counted by character instead, which is what their own tooling does and what a translator quoting per character will bill against. Latin text mixed into the same paragraph is still counted as words, so a bilingual document comes out sensibly.
The full stop that does not end a sentence
"Mr. Patel went to Washington D.C. yesterday." is one sentence. Split on every full stop and you get three — and because average sentence length feeds the readability score, every derived statistic is then wrong too.
Abbreviations, initials, decimals, ellipses and URLs are all recognised before sentences are counted. The abbreviation list is the common set rather than an exhaustive one; getting every case right needs a parser, and getting the frequent ones right needs a list.
Reading ease, and what to do with it
The score is Flesch Reading Ease: 0 is impenetrable, 100 is very easy, and plain English sits around 60 to 70. It is computed from two things only — how long your sentences are, and how many syllables your words have — so it responds to exactly two edits: shorten sentences, and prefer shorter words.
It rests on counting syllables, and English spelling does not encode syllables reliably enough to do that without a dictionary. The heuristic here is the standard one and it is wrong on "queue". That is why the band is shown alongside the number, and why the score is withheld below about 25 words, where a single long word swings it wildly.
Why this one runs in your browser
Consider what actually gets pasted into a word counter. An unpublished manuscript. A dissertation before submission. A job application. A legal filing. A press release under embargo. The common thread is that the text has value precisely because it is not public yet.
Counting words needs no server, so there is no reason to send any of that anywhere. It happens in this tab, and you can watch the Network tab stay empty while you type.
Questions
How do I count the words in my text?
Paste it in. The counts update as you type — words, characters with and without spaces, sentences, paragraphs, lines, reading time and speaking time — and nothing you paste is sent anywhere.
Why does your count differ from Word or Google Docs?
Usually hyphens, contractions or URLs. "Well-known" is one word here and in Word; some counters call it two. A URL is one word here; a counter splitting on punctuation can call it six. If the gap is large and your text is Chinese or Japanese, see below — those scripts are counted by character, which is correct and is what a naive counter gets badly wrong.
How does it count Chinese or Japanese?
By character, because those scripts are written without spaces between words. A page of Chinese has no spaces at all, so a counter that splits on whitespace reports one word for the entire document. Counting characters is what the languages' own tooling does and what a translator quoting per character expects.
What is reading ease?
Flesch Reading Ease, a score from 0 to 100 based on average sentence length and average syllables per word. Plain English sits around 60–70 and most newspapers land near 50. It depends on counting syllables, which in English is a heuristic rather than a rule, so the band is more trustworthy than the number — and it needs about 25 words before it means anything at all.
How is reading time calculated?
At 238 words a minute, which is the median for adult silent reading of prose rather than the round 200 most sites use. Speaking time uses 140, an unhurried presentation pace. Both are averages: technical material reads slower and dialogue faster.
Is my text uploaded?
No, and it matters more here than it looks. The things people paste into a word counter are unpublished manuscripts, job applications, dissertations, legal submissions and press releases under embargo — text whose value depends on it not being anywhere yet. This runs entirely in your browser and makes no request.
Can it show which words I overuse?
Yes — the frequency list is below the counts, with "the", "and" and the rest filtered out by default because a list topped by "the" tells nobody anything. Turn them back on if you want the raw distribution.
What BriskFile will not do to you
-
Your files never leave your device
Every conversion runs in your browser. Open DevTools, watch the Network tab, and you will see no upload — because there is not one.
-
No account, no email, no watermark
Nothing to sign up for and nothing stamped on your images. There is no step between choosing a file and getting it back.
-
No limits at the download button
No daily cap, no file counter, no "upgrade to download". If the tool starts a job, it finishes it.
-
Checkable, not just claimed
Open the Network tab and convert something — nothing goes out. Or load the page, disconnect, and watch it keep working: the tool is already on your machine, and your file never leaves it.