Reading time calculator

How long does it take to read 1,000 words?

At the average adult pace for non-fiction, 238 words per minute, 1,000 words take 4 minutes 12 seconds to read silently. A slower reader at 175 wpm needs 5 minutes 43 seconds; a faster one at 300 wpm is through in 3 minutes 20 seconds. Fiction reads a little quicker, 3 minutes 51 seconds at 260 wpm, and reading the same thousand words aloud takes 5 minutes 28 seconds at 183 wpm.

A few more reference points at 238 wpm: 500 words is just over 2 minutes, 2,000 words is 8 minutes 24 seconds, and 5,000 words is 21 minutes. A 90,000-word novel at the fiction average is about 5 hours 46 minutes of reading, spread over however many evenings it takes. The table under the calculator has the rest, and any row loads into the calculator with one click.

Where the reading speeds come from

Many reading time calculators use a single round number, such as 200, 250 or 300 words per minute. The speeds on this page come from the largest review of the question: Marc Brysbaert’s 2019 meta-analysis in the Journal of Memory and Language, which pooled 190 studies of silent reading with 18,573 participants, and 77 studies of reading aloud with 5,965.

• Non-fiction, read silently: 238 wpm on average, with most adults between 175 and 300.
• Fiction, read silently: 260 wpm on average, with most adults between 200 and 320.
• Reading aloud: 183 wpm on average.

The slower and faster cards in the calculator are the two ends of those ranges, so they describe the spread of real readers rather than an arbitrary margin either side of the average.

The same review explains why the familiar figure of 300 words per minute is too high. It traces back to a few influential sources: early speed tests built on deliberately easy passages, a theory of reading derived from experiments with speeded-up speech, and a much-reprinted table of ten students from one highly selective university. In a book-reading study within the same paper, 48 regular and highly educated readers timed themselves on the next book they read; only 11 of them went faster than 300 wpm. Two caveats apply to all of these numbers: nearly all the participants were educated Western readers, most of them university students, and the averages describe people who stayed on task the whole time.

Matching the estimate to a specific text

Fiction reads faster than non-fiction mostly because its words are shorter. Using word-length counts from very large text collections, Brysbaert found that English fiction averages 4.2 letters per word against 4.6 for non-fiction, and that the difference is enough to explain the gap in reading speed. That led him to suggest a formula for estimating the speed for any particular English text: 238 × 4.6 ÷ the text’s average word length.

The “Match my text” option applies exactly that. It measures the average number of letters per word in what you pasted and adjusts the average speed accordingly, and it scales the slower and faster readers by the same ratio. A text of short everyday words speeds up; a report dense with long technical terms slows down. Brysbaert’s own paper, at 5.1 letters per word, works out at about 215 wpm.

Treat it as an improvement, not a measurement. Word length is only a stand-in for difficulty, as the paper says itself: two texts with the same average word length can read very differently. Checked against an older study with easy and difficult passages, the formula predicted the easy ones almost exactly but still overestimated the speed on the difficult ones, which were read at 182 wpm against a predicted 203.

What changes how long a text takes to read

• Difficulty and familiarity. In that older study, by Britton and colleagues in 1978, the same readers managed 262 wpm on easy passages and 182 wpm on difficult ones. A paper in your own field reads faster than one next door, however long its words are.
• Why you are reading. Reading to remember, as you would for an exam, is far slower than reading to follow. Brysbaert’s review settles on about 100 wpm as a fair estimate for study reading, and several studies found less. For a textbook chapter, type 100 into “Your own speed”.
• Skimming. Skimming is quicker by definition, but it is a different activity rather than a faster version of the same one. A 2016 review of speed reading by Keith Rayner and colleagues found that reading faster comes with lower comprehension, and that what speed-reading courses mostly train is skimming. Skim to find what matters, then read that part at your normal pace.
• Who is reading. Brysbaert reports lower speeds for children, for older adults and for people reading in a second language, where rates below 100 wpm are not unusual. If your audience is mostly reading you in their second language, plan for the slower card or below.
• Screen or paper. Reading times on screen and on paper are now very similar; the early studies that found screens slower used poor displays. Comprehension is where paper still wins. A 2018 meta-analysis by Delgado and colleagues, covering 54 studies, found a small but consistent advantage for paper, larger when readers were short of time and when the text was informational rather than a story.
• Attention. Estimates assume nobody wanders off. In one study students reading a textbook chapter in a lab finished in 34 minutes, while students reading the same chapter at home recorded 1 hour 9 minutes. Messages, tabs and interruptions are not in any formula.

“X min read” labels for blogs and newsletters

A reading-time label under a headline answers the question a reader asks before starting: is this a two-minute read now, or something to save for later? Two blogging platforms that print one say how they work it out:

• Medium says it bases read time on about 265 words per minute, with an adjustment for images that it no longer spells out.
• Ghost, whose code is open source, uses 275 words per minute plus an image allowance, rounds to the nearest minute and never shows less than “1 min read”.
• This calculator uses whichever pace you pick, 238 wpm for non-fiction by default, Ghost’s image allowance, and rounds to the nearest minute with a minimum of one.

Ghost’s image allowance is 12 seconds for the first image, 11 for the second, one second less for each after that, and 3 seconds each from the tenth onwards, so ten images add 75 seconds. Medium’s help centre described the same rule in 2017, when it used 275 words per minute. The allowance suits photographs and illustrations. A chart or a table that has to be studied takes longer, so for a data-heavy report count each one as two or three images, or pick a slower pace.

Because both platforms assume faster readers than the research average, their labels run short: a 1,000-word post is a 4-minute read at 238 wpm and a 3 minute 46 second read at 265. Neither is wrong, but the label is a promise, and a promise that runs over is worse than one that comes in early. A few habits keep labels honest:

• Count the body only: not the title, byline, author bio, footnotes or the list of related posts.
• Use one rule across the whole site, so that a 6-minute read really is twice a 3-minute one.
• In a newsletter, put the label at the top, before the first paragraph, where people decide whether to read now or later.

Reading aloud versus reading silently

Silent reading of non-fiction runs about 30 percent faster than reading aloud: 238 words per minute against 183. Reading aloud is held to the pace at which you can pronounce each word; silent reading is not, and Brysbaert points out that its average is close to the fastest rate at which people can still follow speeded-up speech.

The 183 wpm figure comes from short tests of people reading a passage straight through, so it is a fluent pace rather than a sustainable one. Brysbaert notes that it is unrealistic for someone reading aloud for hours, and cites a small study that considered 120 wpm acceptable across an eight-hour working day. Audiobooks, for comparison, are narrated at roughly 140 to 180 words per minute. Use the read-aloud time for proof-listening a document, recording a chapter, or checking how long it takes to read the minutes out at a meeting.

When to use the speech time calculator instead

If the text is going to be performed rather than read out, reading speed is the wrong measure. A talk, a wedding toast, a video voiceover or a podcast script is delivered to an audience, with pauses, emphasis and the odd laugh, and it runs slower than a fluent read-through. The speech time calculator works at presenting paces of 110, 150 and 190 wpm and also works backwards, from a time slot to the number of words that fits in it. Rehearse the result at a set pace with the teleprompter.

For other kinds of length, the word counter breaks a text down further, the words to pages calculator turns a word count into pages for a given font and spacing, and the sentence counter shows where the long sentences are, which is often why a text reads slower than its word count suggests.

Where VoiceSnap Pro fits

Reading time is the reader’s side of a text. The writer’s side is usually longer, and much of it is typing. Most people speak at around 150 words a minute, faster than most people type, so a 1,000-word first draft is about 7 minutes of talking.

That is the idea behind VoiceSnap Pro, a voice-to-text dictation app for macOS and Windows. Hold one keyboard shortcut, speak, and clean punctuated text appears in whatever field your cursor is already in, whether that is a blog editor, a newsletter draft or an email. Every dictation is also saved to a searchable notes library. It will be a one-time $39 purchase, not a subscription.

It has not shipped yet, so there is nothing to download today. Join the waitlist for one email on release day. In the meantime, the voice typing tool lets you dictate into your browser and see what spoken text looks like on the page.