How this calculator works
Speaking pace turns a script into time through one relationship: words divided by words-per-minute gives minutes. Run it backwards — minutes multiplied by words-per-minute — and a target runtime turns into a word budget. That's the whole engine behind both directions of this tool.
Duration → Words: word count = minutes × wpm
Range: fastest = words ÷ pace-high, slowest = words ÷ pace-low
Word count is counted the same way whether you paste a full script or type a number directly: runs of spaces, tabs and line breaks all collapse to a single gap, so double-spaced scripts, indented dialogue and copy-pasted formatting don't inflate the count. The range shown alongside the headline number isn't padding — it comes from the low and high end of whichever pace preset you picked, so a "conversational" estimate always shows what the same script looks like read a little slower and a little faster.
Where these speaking-pace numbers come from
Conversational English speech in the US runs about 120-150 words per minute, based on National Center for Voice and Speech figures compiled by VirtualSpeech, with 150 wpm commonly used as the general scripted-speech baseline. On camera, creators tend to run faster than that: most YouTubers speak between 140 and 170 wpm, and the exact number shifts with format — tutorials and educational content sit lower at 130-150 wpm so viewers can keep up, vlogs and conversational formats run 140-160, and commentary, essays or tech reviews push 150-170. That's the basis for this tool's three general presets (slow 110-130, conversational 140-160, fast/energetic 170-190) plus the YouTube-explainer preset (135-160).
The other two per-format presets come from their own industries. Audiobook narration is a well-documented 150-160 wpm — steady and unhurried, since listeners sit with it for hours rather than minutes. Commercial voiceover reads naturally at close to 150 wpm, which is exactly why a standard 30-second radio or podcast ad is written to about 75 words; faster retail and promo reads push that toward 175, giving the ad-read preset its 150-175 range.
Why tighter pacing helps YouTube retention
Retention editing leans on cuts roughly every 2-5 seconds to keep the picture moving, and scripts written in shorter sentences make that kind of cutting possible — a long, comma-heavy sentence can't be chopped without losing the thread, but three short ones can. That's part of why high-retention channels can sound like they're talking at 160-190 wpm of perceived pace even though nobody is actually speaking that fast: a meaningful share of it comes from trimming dead air and hesitation out in the edit, not from talking faster on the day. Writing tighter sentences up front — the practice this calculator is built to support — makes both the recording and the edit easier, because there's less silence to cut around in the first place.
There's a comprehension ceiling to respect here too. Comprehension for a native-language listener peaks somewhere around 180-200 wpm for simple material, but speaking above roughly 160 wpm measurably hurts comprehension of complex or unfamiliar content — and dropping under about 100 wpm loses an audience's attention within a couple of minutes. Match your pace preset to how dense the material actually is: a fast-cut reaction video and a technical tutorial shouldn't use the same number.
Worked example 1 — one script, three paces
Take a 600-word script and read it at each of the three general presets:
- Slow (120 wpm): 600 ÷ 120 = 5:00
- Conversational (150 wpm): 600 ÷ 150 = 4:00
- Fast / energetic (180 wpm): 600 ÷ 180 = 3:20
The same 600 words swing by a full 100 seconds — nearly 42% longer at the slow end than the fast end — purely from delivery. That's why "how many words for a 5-minute video" doesn't have one answer: it depends entirely on who's reading it and how.
Worked example 2 — a fixed target, reversed
Now flip it: you've been told the video must land at 8:00 exactly, and you need to know how much to write.
- Slow (120 wpm): 8 × 120 = 960 words
- Conversational (150 wpm): 8 × 150 = 1,200 words
- Fast / energetic (180 wpm): 8 × 180 = 1,440 words
A 480-word gap between the slow and fast budgets for the identical runtime — which is exactly the reverse-mode calculation above, useful before you've written a word rather than after.
Pacing tips
- Budget 10-15% extra time for what isn't speech. Breaths, beat changes, b-roll cutaways, on-screen text holds and edited pauses routinely add that much on top of a pure words-÷-wpm estimate, so treat this tool's number as a floor, not a guarantee, especially on longer videos.
- Write shorter sentences if you want a faster-feeling video. Short sentences are easier to cut around, which is most of how "fast-paced" editing actually gets made — not by talking quicker, but by leaving less silence to trim.
- Match pace to density, not habit. A number-heavy tutorial earns a slower pace than a reaction video, even from the same creator in the same week.
- Re-time after every edit pass. A script that reads at 4:00 in your head often runs longer once you actually record it — your real pace under a mic drifts from your imagined one, so time a real take once and adjust the wpm field to match your own number.
- Ad and audiobook work has less pacing freedom. A 30-second spot needs to land in 30 seconds; write to the word count the slot allows rather than discovering the mismatch at the read.
All figures here are general industry ranges, not a measurement of your own voice. Time yourself once at your actual pace and use that number for anything that has to hit an exact runtime.
FAQ
How many words per minute is normal for talking?
Conversational English speech in the US runs about 120-150 words per minute, per the National Center for Voice and Speech figures compiled by VirtualSpeech, with 150 wpm often used as the general baseline for scripted delivery. On camera, most YouTubers speak faster than that — typically 140-170 wpm — because a camera performance is usually more energetic than a hallway conversation.
What speaking pace should I use for a YouTube video?
It depends on the format. Tutorials and educational explainers tend to sit at 130-150 wpm so viewers can follow along, vlogs and conversational content run 140-160 wpm, and commentary, essay-style or tech-review videos push 150-170 wpm, per Prepublish's breakdown. High-retention channels often sound even faster than that — 160-190 wpm of perceived pace — but a chunk of that comes from cutting out pauses in the edit, not from talking faster on the day.
Why does my video always run longer than my script predicts?
Because a raw words-divided-by-pace calculation only counts the words you say — it does not count the seconds you spend not talking. Breaths, beat changes, b-roll cutaways, on-screen text holds and edited pauses commonly add another 10-15% on top of the spoken-word estimate, which is why experienced scriptwriters build that margin in rather than trusting the math to the second.
Is there a standard audiobook narration speed?
Professional audiobook narration typically runs 150-160 words per minute — slow enough to sound natural and unhurried over hours of listening, but not so slow that it drags. It's noticeably slower than a fast-paced YouTube commentary track and close to ordinary conversational speech.
How fast do ads get read?
A natural, conversational ad read lands close to 150 wpm — which is why a standard 30-second radio or podcast spot is written to about 75 words. Energetic retail or promo reads run faster than that, and slower, emotional or luxury-brand reads run well under it, so 150-175 wpm is a reasonable working range unless the brief says otherwise.
Does a faster pace hurt viewer comprehension?
Past a point, yes. Comprehension for a native-language listener peaks somewhere around 180-200 wpm for simple material, per VirtualSpeech's summary of the research, but speaking above roughly 160 wpm measurably reduces comprehension of complex or unfamiliar content, and going below about 100 wpm tends to lose an audience's attention within a couple of minutes. Match the pace to how dense the material is, not to a single target number.