Workman typing test
Loading the test…
Practise the Workman keyboard layout here without reinstalling anything. The test emulates Workman in the browser, translating each physical key you press into the letter Workman puts there, so you can type on a QWERTY laptop and be scored as a Workman typist. You'll get net WPM, accuracy, a per-key breakdown and, when you're signed in, a Workman personal best that stays separate from every other layout you test.
Workman at a glance
OJ Bucao published Workman on 6 September 2010, Labor Day in the US, and named it for people who type for a living. His write-up objects to one thing in Colemak above all: the centre columns. Reaching sideways with the index finger for H, which Colemak puts in the centre column, forces a lateral wrist movement on HE, one of English's commonest pairs. Workman was built around a different model of finger effort, one where reaching straight up or curling down is cheap and stretching sideways is expensive.
Here is the full letter layout, row by row, as published in the project's own key map:
| Row | Left hand | Right hand |
|---|---|---|
| Top | Q D R W B | J F U P ; |
| Home | A S H T G | Y N E O I ' |
| Bottom | Z X M C V | K L , . / |
A few things stand out. The home row holds S, H, T, N, E, O and I, and A stays where it always was. The two centre columns carry B, G, V, J, Y and K, mostly rare letters, so the inner stretches happen less often. According to the Workman site, only 6% of keystrokes land on those middle columns, compared with 12% on Colemak.
The comma, period, slash and apostrophe stay on their QWERTY keys; the semicolon moves up to where P was. Z and X don't move. C and V each shift one key to the right, so undo, cut, copy and paste still sit under one hand, but copy and paste need a small adjustment.
How many keys change
Workman's designer counts 21 characters that differ from QWERTY, against 31 for Dvorak. Colemak moves fewer, 17 keys, which makes it the smaller change from QWERTY. Workman asks you to relearn a few more positions in exchange for its approach to the centre columns.
| Layout | Keys changed from QWERTY | Source of the count |
|---|---|---|
| Colemak | 17 | Colemak |
| Workman | 21 | Workman |
| Dvorak | 31 | Workman |
The counts come from each project's own documentation, so treat them as rough rather than as a single consistent measure.
Using the emulation
The test reads which physical key you pressed and swaps in the Workman character for that spot. Press the key printed D on QWERTY keycaps and you'll type H. Your operating system never changes, so this is a cheap way to find out whether Workman suits your hands before you commit to it.
Already typing Workman system-wide? That's fine here: emulation reads positions, not the letters your system produces, so you get the same result. Keyboard firmware is different, because it moves the keys before the browser sees them, so each keypress would be converted twice; emulation can't be switched off on this page, so use tests elsewhere with layout emulation off in settings. The same goes if you use a screen reader that announces keys from your system layout.
While you learn, the keymap below the text shows where the next letter lives on Workman.
Where Workman learners get stuck
The hardest part is that some letters you never think about move a long way. D jumps from the home row to the top row, and E moves from the left hand to the right. On QWERTY, "the" alternates hands; on Workman, T and H are adjacent on the left home row and E sits on the right. These patterns take time to rewire, so give them dedicated practice rather than hoping they settle on their own.
A short routine that works with this page:
- Run three 1-minute tests on QWERTY first and note the median. That's the number Workman has to beat.
- Choose Workman under layout emulation in settings, then start with 10-word tests.
- After each run, read the most-missed keys on the results screen. Write down the two worst.
- Use practice to drill those keys before your next test.
Keep accuracy above about 95% before pushing for speed. Net WPM gives you nothing for a word with one wrong letter, so a fast, sloppy run tells you less about your progress than a slow, clean one.
Workman or Colemak-DH?
Both answer the same complaint about standard Colemak, and they answer it differently. Workman starts again from QWERTY with its own model of finger cost. Colemak-DH keeps Colemak's home-row vowels and moves only D, G, B, H and M. If you're starting from QWERTY with no alternative layout yet, try a week of each with emulation and keep the one that feels better after a long session. The Colemak-DH typing test keeps its own personal best in the same way, so the comparison is easy to read.
The broader evidence on whether any alternative layout makes you faster is mixed, and our post on QWERTY vs Colemak vs Dvorak goes through it.
Frequently asked questions
Can I try Workman without installing it?
That's what this page does. Keep emulation on, type on your normal keyboard and watch the keymap for the new positions.
Does Workman keep keyboard shortcuts usable?
Mostly. Z and X stay put, and C and V shift one key right, so undo, cut, copy and paste are all still reachable with the left hand, as the Workman site points out.
Is there a Workman variant for programmers?
The project mentions Workman-P, which swaps the number row so symbols are typed without Shift. The test here emulates the standard Workman layout shown in the table.
What's a good Workman typing speed?
Compare it with your own QWERTY median. Matching it is the milestone most switchers aim for, and for context, the average in a 2018 Aalto University and Cambridge study of 168,000 volunteers was 52 WPM.
How is the score calculated?
Net WPM counts the characters of fully correct words, five characters to a word. See the scoring methodology for the formula.