What is WPM, and how this site calculates it
WPM means words per minute. In typing measurement, the word is not a dictionary word. It is a block of five characters, counting letters, spaces, and punctuation the same way. A short word and a long word would otherwise make the same pair of hands look fast or slow depending on the passage. Five characters is the usual way to stop that. The home page says this briefly. This page is the one to link when you need the version that matches the code on the practice screen.
The practice-test formula
On the Practice screen, live WPM is the number of correct characters so far, divided by five, divided by the minutes since your first keystroke, then rounded to the nearest whole number. If no time has passed, the display stays at zero rather than dividing by nothing. Accuracy is correct keystrokes divided by all keystrokes, rounded to a percentage. Before you have typed anything, accuracy shows as 100, because there is nothing to get wrong yet.
A wrong key is painted on the character and the cursor advances. You are not asked to backspace. That wrong key increases the error count and the total keystroke count. It does not increase the correct-character count, so it cannot raise WPM. It does lower accuracy. There is no second penalty that subtracts the error again from the word count. If you have seen a formula that takes characters, divides by five, then subtracts the error count, that is a different test. Do not line the two numbers up and call the gap a change in your skill.
Here is a small example you can check with a calculator. Suppose a 30-second Timed run records 150 correct characters and 10 wrong ones. Words are 150 divided by 5, which is 30. Minutes are 30 seconds divided by 60, which is 0.5. WPM is 30 divided by 0.5, which is 60. Accuracy is 150 correct divided by 160 total keystrokes, which is about 94 percent. The results panel would show 60 WPM and 94 percent accuracy. The same arithmetic runs while you type. It is not a separate “official” score applied only at the end.
When the clock starts and stops
The timer and the WPM clock both wait for the first character. Staring at the passage does not start them. In Timed mode the test ends when the countdown reaches zero. You chose that countdown before you started: 15 seconds, 30 seconds, 60 seconds, or 2 minutes. In Words mode there is no countdown. The test ends when you reach the end of the generated passage, and WPM uses however many minutes that took. Same formula, different stop rule. That is why a Words score and a 15-second score can both be “true” and still disagree. The short test is a smaller sample.
The Code tab still updates the live counters from correct and total keystrokes. It does not run the Timed countdown, and it does not pop the end-of-passage results panel the way Words mode does. If you want a code-flavored score you can save, set the level menu to Code while you are in Timed or Words. That menu feeds programming keywords. The Code tab feeds short lines of actual syntax. Both are described in the practice guide.
Lessons use the same shape of formula on their own counters: correct characters divided by five, divided by minutes, and accuracy as correct divided by typed. A lesson score is for that attempt. It is not written into rankings. Type Racer also shows a WPM from correct characters, but a wrong character does not advance the passage. You stay on the letter until you type it. A Racer WPM is therefore not the same event as a practice-test WPM, even when the five-character idea is shared. The games guide says this again next to the ghost, which stores a Racer speed in this browser and nobody else’s.
Bands, as labels rather than judgments
These ranges are a reading aid. They are not a placement exam, and they are not a ranking of every person who has opened the site.
| WPM | Label | What to do with it here |
|---|---|---|
| 1–30 | Beginner | Stay on home row and Easy words. Do not open Hard. |
| 31–60 | Average | Sixty-second tests. Move rows only when accuracy holds. |
| 61–90 | Proficient | Medium text, occasional two-minute runs, games as variety. |
| 91–120 | Fast | Hard words or Code if that is your real work. Watch the second minute. |
| 120+ | Elite | Small, specific weaknesses. A ghost in Type Racer is a better target than a new gimmick. |
A job posting that says “60 WPM” is talking about whatever test that employer uses. It may require corrections. It may use a set passage. It may ignore accuracy or enforce it strictly. Practice here until your 60-second runs sit comfortably above the number with accuracy near 95 percent, and treat that as preparation, not as the employer’s score. The page on why speed matters goes further into that limit. Nothing on SumnuTyping is an official certification.
What people misread
The characters figure on the results panel is the correct-character count, not the length of the passage and not the count of mistakes. Errors are their own box. Personal best on this browser updates when a finished test beats the stored best WPM. It does not update from a game, and it does not update from a lesson. Rankings, if you are signed in and you choose to post, are a separate act. The board you see keeps a strong score per name in the set it loaded. It is not a second definition of WPM.
Compare yourself with yourself, on the same mode and the same duration, a week apart. That is the only comparison this formula was built to support. Open a 60-second test when you want a sample, or go back to the practice plan if the number is moving for the wrong reason.