Reading time of a text is calculated by dividing its word count by a reading rate measured in words per minute, so a 1,000-word article read at the average adult silent rate of 238 words per minute takes roughly 4 minutes 12 seconds. That single division explains every reading-time estimate you have ever seen, but the numbers that come out of it depend on two inputs you have to choose deliberately: the rate you assume and what you count as a word. Silent reading and reading aloud are different tasks with different measured speeds, which is why a serious tool reports both side by side instead of forcing one figure to cover two situations. The Reading Time Calculator follows that approach: paste text or type a word count, pick a preset for each activity or override the rate, and read both durations next to the exact word count used in the arithmetic. Because the inputs are visible and adjustable, the result is something you can defend when a stakeholder asks where the number came from, instead of a magic figure from a black box.
How to calculate reading time of text with the Reading Time Calculator
The Calculator runs entirely in the browser, so nothing you paste is uploaded or saved, and the same workflow covers every text source from blog drafts to speech scripts.
- Paste your text into the input box, or switch to word-count mode and type the number of words directly if you only have a count and not the source.
- Pick a preset reading rate for silent reading, or type your own words-per-minute value, and set a separate speaking rate for the read-aloud estimate.
- Read both durations side by side, and check the word count and counting rule printed under the result so you can audit the arithmetic by hand.
If the count surprises you, the rule under the result tells you why: text is split on whitespace after word-joining dashes are treated as separators, and a token counts as a word only when it contains at least one letter or digit. That is why a hyphenated compound like state-of-the-art counts as one word, a contraction counts as one, a number like 3.14 or 1,000 counts as one, and a string of pure punctuation counts as none. If you are pulling text out of a Word document, the same rule applies once you paste it in; you can also consult a guide on how to calculate reading time in Word documents for the document-specific steps.
Default rates and where they come from
The default rates on the Calculator are not folklore; they are pinned to published measurements so you know what population your estimate is averaging over.
| Activity | Default rate (wpm) | Source |
|---|---|---|
| Silent reading, non-fiction (English) | 238 | Brysbaert 2019 meta-analysis (190 studies) |
| Silent reading, fiction (English) | 260 | Brysbaert 2019 meta-analysis |
| Reading aloud (oral reading) | 183 | Brysbaert 2019 meta-analysis (77 oral-reading studies) |
| IReST standardized passage reading | 228 | Trauzettel-Klosinski & Dietz 2012 (IReST study) |
| Presentation speaking | ~150 | Convention used by speech coaches; no single retrievable study asserts this exact figure |
The Brysbaert 2019 meta-analysis is the single largest source behind these defaults: 190 studies of silent reading and 77 studies of oral reading, summarized in the Journal of Memory and Language. The IReST figure is offered alongside the oral-reading preset because it measured a specific kind of task, reading short standardized paragraphs under a stopwatch, which is faster than reading naturally but slower than silent reading. The familiar presentation number of 150 wpm is honestly labelled as a convention rather than a measured average, because no retrievable study actually pins it down; treat it as a planning lower bound, not a population mean.
Silent vs spoken estimates: when each one matters
The reason the Calculator reports both durations is that "how long will this take?" means two different things in two different workflows. Pick the one that matches the situation you are budgeting for.
| Situation | Use the silent estimate | Use the spoken estimate |
|---|---|---|
| Blog post or article preview | Yes — readers will skim at their own pace | No — most readers will not read aloud |
| Conference talk or keynote | No — pacing is dictated by your delivery | Yes — base the slot on the read-aloud time |
| Podcast intro or script | No — every word is spoken | Yes — pad for pauses and breath |
| Email or short newsletter | Yes — recipients will scan silently | No — unless you expect screen-reader use |
| Tutorial with embedded code | Yes — readers stop to copy and run | No — speaking time is rarely relevant |
| Voiceover or narration | No — timing must include breath and pacing | Yes — and add a margin for retakes |
The split is not pedantic. A 10-minute talk paced at 150 wpm holds roughly 1,500 spoken words; the same script read silently at 238 wpm would only take about 6 minutes 18 seconds, a gap that is the difference between a comfortable keynote and an over-stuffed one. Use the spoken rate whenever the audience is hearing the words, and the silent rate whenever they are choosing to read them.
Why reading time tools disagree with each other
If you paste the same text into three different calculators and get three different minutes, the disagreement almost always traces back to one of two places: the counting rule or the assumed rate. A token that one tool counts as a word, another splits in two or three on hyphens, apostrophes, or decimal points. Hyphenated compounds like state-of-the-art are the most common flashpoint: counted as one word, four words, or three depending on the rule. Once the counts differ, even identical rates give different minutes.
The second source of disagreement is the rate. A tool that defaults to 200 wpm will report a longer time than one that defaults to 260 wpm, even on the same word count, and neither is wrong against its own assumption. The Calculator exposes both inputs so you can see why: the word count and the counting rule are printed under the result, and the rates are shown next to each duration. If a number looks off, change one input at a time and watch the result move; that is the only honest way to compare estimates between tools.
Planning with reading time estimates
Reading time figures are population averages, not promises, and the Calculator is precise about its arithmetic without pretending it can predict a single reader. Brysbaert's own review places most adults between 175 and 300 words per minute silently, which means the true reading time for a particular person can differ from the estimate by roughly a third in either direction. Treat the result as a planning figure: useful for comparing drafts against each other, useful for sizing a presentation slot, useful for telling a reader in a preview that an article is "about a 7 minute read", and not useful as a guaranteed number of minutes.
A few practical habits make the estimate more useful in real work. First, compare drafts at the same rate so the only thing changing is the word count; this tells you whether an edit actually shortened the time, or just shuffled it. Second, when budgeting a presentation, add a margin of 10 to 15 percent on top of the read-aloud estimate to absorb pauses, audience reactions, and the occasional stumble, because no human delivers at the rate a stopwatch measures. Third, when telling a reader what to expect, round up rather than down, since under-promising keeps the trust. And finally, if the word count on the page does not match your own count, trust the rule printed under the result over your gut, because the rule is the reason.
For a deeper look, see Reverse Text in PowerPoint: A Copy-Paste Workflow.