Speaking Speed: How Many Words per Minute People Talk

People speak at different rates in ordinary conversation, and the exact figure depends on whether you count the pauses. In a large corpus of recorded conversations, the overall rate, counting all elapsed time, was 196 words per minute ¹. Remove the silences and non-speech noise, and the net rate rises to 236 words per minute ². Silent reading of non-fiction is faster still, at about 238 words per minute for adult readers of English ³, and reading aloud is slower, at about 183 words per minute ⁴.

This guide sets out those rates side by side, with the study behind each one. It belongs to the guide to typing speed, and it pairs with the average typing speed and the speaking vs typing comparison. Short answers to the most common versions of the question are linked in each section.

Ordinary conversation averages 196 words per minute with pauses counted, and 236 without them

The most useful single number for everyday speech is the overall rate in conversation. Researchers Jiahong Yuan, Mark Liberman and Christopher Cieri added up the words spoken by both speakers in a corpus of conversational telephone speech, in English and Chinese, and divided by the elapsed time of each call. The average was 196 words per minute ¹. That is the rate you would hear if you timed a real conversation with a stopwatch.

Individual conversations vary widely. Measured over elapsed time, the rate ranged from 111 words per minute to 291 words per minute across conversations ⁵. That spread is between conversations, so it does not show how individual speakers compare with one another.

The corpus is spontaneous telephone speech. It is not dictation, a presentation or a formal interview, so read the table as a measure of relaxed talk.

Speaking rate counts pauses and articulation rate does not, which is why reported speeds differ

Much of the gap between reported speeds comes down to what is counted. The overall rate divides words by all the time a conversation takes, so every pause, hesitation and turn-taking gap lowers the number. The net rate, which researchers also call the articulation rate, counts only the time in which words are actually being spoken. The net rate in the same corpus was 236 words per minute ².

The gap between the two is large. Taking the overall rate from the net rate shows that about 17 percent of elapsed conversation time carries no speech at all ⁶. A speaker who says "about 150 words per minute" may be counting pauses, may be reading from a script, or may be using a different measure altogether. Before you compare two numbers, check which one they measure. Our page on the articulation rate explains the term in more detail.

Each measure answers a different question. The overall rate tells you how long a recording or a conversation will take, which is what a planner needs. The net rate tells you how fast a voice moves when it is moving, which is closer to the question of how fast a speaker can articulate. Both are real, and a figure quoted without its definition is hard to use.

The rate also changes with the length of each turn. In the same corpus, speaking rate rises rapidly as a speaker's turn grows from one word to seven words ⁸. A measure taken over short replies such as "yes" or "right" will therefore read lower than one taken over longer turns.

Presentations are a separate case, and the measured evidence is thin

Presentations are not the same as conversation, and the studies that compare them are few. The clearest measured comparison looked at TED talks and university lectures. Peter Wingrove sampled one-minute stretches of both and found that speech was significantly faster in TED talks than in lectures, measured in syllables per second ⁹. The study does not say that presenters talk faster than conversation, and the abstract does not give the mean speeds, so we do not quote a lecture figure from it.

So the honest answer to "how many words per minute should I speak in a presentation?" is that no large measured study fixes one number for presenters. Coaching advice on presentation pace is common, but we found no measured study behind a single target. The better test is your own recording, timed as described at the end of this guide. The question on how many words per minute to speak in a presentation takes the advice in more detail.

A simple planning check shows why the overall rate matters. At the conversational average of 196 words per minute ¹, a script for a talk is sized by multiplying the minutes by that rate. The rate already includes the ordinary pauses of conversation, so slides, applause and questions add time on top of it, and a script written to that count will usually run long. A script written to the net rate of 236 words per minute ² will run even longer in practice.

The overall rate of a talk depends on what the speaker does besides talking. A speaker who stops to let a slide land, waits for a laugh or reads a quotation aloud lowers the overall rate without speaking any more slowly. That is the same effect the pause measure shows in conversation, and it means a talk and a chat can have similar spoken rates but very different overall rates. When you set a pace for a presentation, decide first whether you mean the rate while talking or the rate across the whole slot.

The Wingrove study was framed around academic listening: it asks whether TED talks can stand in for lectures in courses that train students to follow academic speech. Its finding is about how the two genres are delivered, not about how well either one is presented.

Audiobook narration is slower than conversation, at a typical 140 to 180 words per minute

For narration, the figure we found in the research literature comes from Marc Brysbaert's 2019 review of reading rate. While arguing that reading can run faster than listening, he writes that audiobooks are spoken at a rate of 140 to 180 words per minute ¹⁰. He uses 150 words per minute as the working speech rate in the same argument ¹¹, which sits inside that range and below the conversational average.

That range is a quoted estimate, not a survey of narrators, so treat it as typical rather than exact. Narrators vary with the genre, the energy of the text and the listener's settings. The question on how many words per minute for a voice-over or audiobook covers the production side, including why a narrator's pace is not the same as the listener's playback speed.

The fastest recorded speech is a short burst, and sustained fast speech is found in rap

The fastest recorded talker in English holds a record for a short, memorized text. Sean Shannon recited the 260-word "To be or not to be" soliloquy, and Guinness World Records lists the rate as 655 words per minute ¹². That is about 3.3 times the conversational average of 196 ¹³. A speed of that kind is a burst over a memorized text, not a pace for a talk.

Sustained speech at high speed is found in rap. Eminem's "Rap God" contains 1,560 words and runs 6 minutes 4 seconds, according to Guinness ¹⁴. Averaged over the whole track, that is about 257 words per minute ¹⁵. Because the track is averaged over its full length, it is a better measure of sustained fast delivery than the short record, although rap is performed to a beat and is not a model for unscripted speech. The question on the fastest speaking speed lists the other records people ask about.

Speed has a listening cost. In a study of the ability to catch speech recognition errors by ear, researchers found that the ability degraded at higher speech rates, and they tested at 300 words per minute ¹⁶. Fast speech can be followed, but a listener's ability to notice mistakes in it drops as the rate climbs. That is one reason record-breaking speech is a feat of practice rather than a model for a talk.

Silent reading of non-fiction is faster than conversation, and reading aloud is slower

Silent reading is faster than conversation. Brysbaert's meta-analysis of 190 studies with 18,573 participants estimates the average silent reading rate for adults in English at 238 words per minute for non-fiction, and 260 for fiction ³. Relative to the conversational average, silent reading of non-fiction is about 1.2 times as fast ¹⁷.

Reading aloud is slower. Across 77 studies with 5,965 participants, the average oral reading rate was 183 words per minute ⁴. Reading aloud has to produce each word, which is why it sits closer to the speaking rate than silent reading does.

Averages also hide who is reading. Brysbaert reports that reading rates are lower for children, for older adults and for readers who use English as a second language ¹⁸. A single adult average is therefore a benchmark for adults, not a target for every reader.

Some claims about reading go much higher. Rayner and colleagues, in a 2016 review, concluded that claims of reading at super speed with good comprehension are overstated and not in line with what research shows about how language is processed ¹⁹. That covers the "speed reading" programs that promise much faster reading than the average adult achieves. The question on whether you can read 900 words per minute explains the evidence, and the question on average reading speed gives the adult figures in more detail. For children, see reading speed by grade, and for the measure itself see oral reading fluency.

In one controlled phone study, speech was nearly three times faster than the keyboard

When the goal is raw words on the page, speech can be much faster than typing. In a controlled study on an iPhone, English text entered by speech reached 153 words per minute, against 52 words per minute on the built-in keyboard, 2.93 times faster ^{7 20}. The speed counted the finished text, including time spent correcting errors, but the phrases were short and were read rather than composed. The same study found a slightly smaller gap in Mandarin Chinese, where speech reached 123 words per minute against 43 on the Pinyin keyboard, 2.87 times faster ^{21 22}. Language and input method can change the size of the gap, so one ratio should not be carried over to every writer.

Larger studies of typing give similar or lower figures. A study of 136 million keystrokes from an online typing test found an average of 51.56 words per minute on a computer keyboard ²³. Those volunteers chose to take the test, so the average is likely higher than the figure for all typists. On phones, an online study of 37,370 volunteers who typed on their own devices found an average of 36.2 words per minute ²⁴. The chart places all of these numbers together.

The chart mixes studies with different tasks, so read the bars as an order of magnitude, not as one test. The gap that matters for writing is the one between speech and the keyboard. Dictation is faster only for the first draft. Dictated words need checking and correction, and a dictated draft takes its own editing time. Our guide on dictation vs typing covers that trade-off, and Yap, which we make, is an on-device dictation app for Mac and Windows, so it is one way to try speaking your text. The keyboard limit itself is covered in the guide to how to type faster.

Measure your own rate: words spoken divided by minutes, from a recording

The published averages describe groups. To learn your own rate, time a sample of speech and count it the same way the research does.

1. Record yourself speaking for at least two minutes on a topic you know well, such as describing your job. Use a phone or any recorder, and keep the mic at the same distance you would use in a meeting.
2. Note the length of the recording. Use the total elapsed time, including pauses, to get the overall rate. The net rate needs the time with speech removed, which takes more effort to measure by hand.
3. Turn the recording into text. A free option is audio to text in your browser, which produces a transcript you can count.
4. Count the words in the transcript. Most word processors show a word count after you paste the text.
5. Divide the word count by the minutes of the recording. For example, 390 words in 2 minutes is 195 words per minute, close to the conversational average.

Repeat the test in two settings, such as a casual chat and a short talk. The gap between the two is usually more useful than either number alone, and it shows how much your pace changes with the setting. For the keyboard side of the comparison, see the guide to how typing speed is measured.