Reading Time & Speaking Time Calculator
Estimate how long it takes to read or speak a text — adjustable WPM presets for slow, average, fast readers, TED, podcast.
Presets
Reading time
—
Speaking time
—
Words
0
Characters
0
Sentences
0
Paragraphs
0
Estimates assume silent reading and continuous speaking — pauses, slides, and audience reaction will add time.
How to use
- Paste your text — article, blog post, talk script, anything.
- Adjust reading and speaking WPM, or pick a preset.
- Times update live alongside word/sentence/paragraph counts.
About this tool
ShowHide
Paste text and get an estimated reading time along with the counts behind it — characters with and without spaces, words, sentences and paragraphs — at a words-per-minute rate you can set for silent reading or for speaking aloud. Mixed-language text is handled properly rather than by splitting on spaces alone: Latin text is counted in whitespace-separated tokens, while each Chinese, Japanese or Korean character counts as one unit, because those scripts do not put spaces between words. That means a CJK figure is closer to a character count than a word count, so lower the rate for those languages — the usual advice is a few hundred characters per minute rather than a few hundred words. Everything runs locally in your browser.
Frequently asked questions
Where does 238 wpm come from?
Brysbaert (2019) meta-analyzed 190 studies and found 238 wpm for silent reading of English non-fiction. 175 for fiction, 300+ for skimming. 238 is a good default.
Is 140 wpm right for a presentation?
140 is conversational; presentations often slow to 100–130 wpm for emphasis. TED talks average ~155 wpm. Pick the preset that matches your style.
How is CJK counted?
Each CJK ideograph (Chinese hanzi, Japanese kanji/kana, Korean hangul) counts as one 'word' for both reading and speaking — close to the right estimate without needing language-specific tokenizers.
Why no 'characters' time?
Reading speed correlates with words, not characters. Variation in word length averages out in any reasonable sample.
Why is the estimate for Korean or Japanese text so long?
Because each hangul syllable, kanji and kana counts as one 'word', which is not the unit the wpm figure was measured in. A four-word Korean sentence built from eleven syllable blocks is counted as eleven, so the total runs roughly 2.5-3× high and the time comes out that much too long. Raise the reading speed by the same factor to compensate — if your text averages three syllables per word, set 700 rather than 238 — or time yourself on a hundred syllables once and reuse that number.
How are sentences and paragraphs counted?
A sentence ends at one or more of . ! ? 。 ! ?, so CJK full stops are handled but a text with no terminal punctuation at all still counts as one sentence rather than zero. Abbreviations inflate the count — 'e.g.' reads as two sentence ends — and a decimal point does the same. Paragraphs are runs separated by a blank line, so a document that uses single line breaks between paragraphs reports one paragraph. Both figures are for orientation, not for a style checker.
Related tools
Typing Speed Test (WPM)
Measure your typing speed in WPM and CPM with accuracy tracking. 15/30/60/120s tests, English / Korean / Japanese word lists, personal best persists.
Text
Bionic Reading Converter
Convert any text into bionic reading format — first letters of each word bolded as fixation points to help your eyes glide faster through long passages.
Text
Markdown Stats & Reading Time
Count words, sentences, headings, links, code blocks, and estimate reading / speaking time for any Markdown.
Text
Bandwidth & Download Time Calculator
Estimate download time from file size and connection speed — or the speed required to transfer a given size within a deadline.
Network
Readability Calculator (Flesch-Kincaid, SMOG)
Score any English text on six classic readability formulas — Flesch Reading Ease, Flesch-Kincaid, Gunning Fog, SMOG, ARI and Coleman-Liau.
Text
Character & Word Counter
Count characters, words, sentences, lines, and bytes in real time.
Text