Typing Speed Statistics 2026: Live Data from Real Tests

Why live data instead of a frozen chart

Most “typing speed statistics” pages repeat the same handful of figures without saying where they came from. This one shows its working: the numbers below are counted from our database when you open the page, using count queries that never read an individual result or name. If the site gets busier, the sample grows and the figures update with it.

Methodology

• Unit of analysis: one completed test by a signed-in user. Guests are not counted.
• Metric: net WPM — correctly typed characters ÷ 5 ÷ minutes (see methodology).
• Bands: 0–19, 20–29, 30–39 … 100–119, 120–149 and 150+ WPM.
• Median band: the band containing the middle test when all tests are ordered by WPM.
• Mixed test lengths: 15-second to 10-minute tests are pooled. Short tests run faster, so the distribution skews slightly high.
• Self-selection: people who practise typing are faster than the general population. Treat these as “typing-site users”, not “all adults”.

What we do not have — and won’t make up

Many statistics pages break typing speed down by country, age and device. We do not collect any of those for tests, so we cannot publish them, and we would rather say so than invent them. For population-level figures, the best source is the Aalto University and University of Cambridge study of 136 million keystrokes from roughly 168,000 volunteers, which measured an average of about 52 WPM — itself from people who chose to take an online typing test.

How to read the thresholds

The “share at or above” rows line up with common job requirements: 40 WPM is the US federal and many state clerical standards, 50–60 WPM is typical for secretaries and chat agents, and 70 WPM is a common bar for transcription. See every requirement on the requirements table, and check your own position with the percentile tool.

Citing this report

You are welcome to use these figures with credit: “Ai-TypingTest.com, Typing Speed Statistics (live), accessed [date]”, linking to this page. Please quote the sample size shown at the time you accessed it, and the self-selection caveat above.

Why typing-site averages run higher than “average typing speed”

Anyone who chooses to take a typing test is, on average, a keener typist than the general population. That is true of this site, of every other typing site, and of the large published studies, which recruited volunteers online. The figure most often quoted for adults — around 40 WPM — comes from older office-skills surveys and is best read as a rough population estimate. The Aalto University study, with 136 million keystrokes, found a mean of about 52 WPM among its volunteers, with the fastest typists above 120 WPM.

Three other things push typing-site numbers up or down:

• Test length. 15- and 30-second tests produce higher WPM than five-minute tests; we pool all lengths.
• Text. Easy, common words are faster than business prose with capitals, numbers and punctuation.
• Repeat testers. Regular users take many tests, so they are over-represented in a per-test count.

What the bands mean for job seekers

Reproducing this analysis

The distribution is computed with one count query per band boundary over the results collection: the number of tests at or above each boundary, differenced to get each band. The median band is the first band where the running total passes half the sample. The CSV you download contains exactly the numbers shown, with the access date and sample size in a footer line so a citation can be checked later.

Limitations

• No demographic or device breakdown — we don’t collect it for tests.
• Only signed-in tests are counted; guests are not recorded.
• Results are self-recorded in the browser, as on any typing site; extreme outliers are possible.
• The sample changes daily, so quote the date and sample size with any figure.

For your own position in the distribution, use the percentile tool; to see what your speed qualifies you for, use the requirement checker.

Loading live data…

How we handle numbers

Every figure on this site is either quoted from a source we checked, computed from our own data with the method stated, or labelled as a typical range. Sources show the date they were checked; requirement tables are reviewed quarterly; and we never present a typical range as an official rule. If you spot an out-of-date figure, tell us.

You are welcome to cite our tables and statistics with credit to Ai-TypingTest.com and a link. Formulas for WPM, accuracy, KPH and pass rules are on the methodology page.

Sources

1. Observations on typing from 136 million keystrokes (Dhakal et al., CHI 2018) — Aalto University (checked October 18, 2026)