A quick guide.
Everything you need to know about practice and your data.
Which exercise is right for you?
- Learning path
- A placement test checks five areas. Then you practise one lesson at a time.
- Adaptive
- Keys and combinations you find difficult appear more often.
- Finger paths
- One target key in four steps: find, switch, words, test without help.
- More exercises
- Custom texts, transpositions, longer texts or individual weak keys.
Typing, correcting & pausing
After an error, the character stays in place. Simply type the correct key. Backspace lets you go back; previous errors remain counted.
In the learning path and finger paths, F2 reports a different finger for the last keystroke. In finger paths, you repeat that keystroke. The keyboard cannot detect which finger you use.
After each guided section, confirm which fingers you used. Only then does Space continue. Escape pauses and Enter resumes. Leaving an exercise pauses it automatically.
The hand guide shows the key and finger. It starts fading after 12 correct keystrokes and disappears after 40. An error brings it back. Tests start without help.
Unlocking & follow-up tests
A lesson is confirmed after two consecutive accurate unlock rounds: at least 60 non-space characters, your accuracy goal overall and on the first attempt for each character, confirmed finger choice, no reported deviations and no hand guide.
A less accurate or unfinished round only resets this streak. Unlocked lessons stay open. Guided practice does not count towards unlocking. Learning path progress is saved separately for each keyboard layout.
The complete placement test can provisionally skip consecutive lessons you already handle well. In finger paths, the four steps are followed by a follow-up test after at least 24 hours using the same text. The suggested ten-second break is optional.
Good rounds do not prove lasting learning. The sequence, thresholds and intervals are product rules, not a scientifically optimised learning formula.
Accuracy, speed & pulse
Accuracy is the proportion of correct keystrokes across all attempts. Unlocking also considers whether each character was correct on the first attempt.
Words per minute only counts complete words typed without errors or corrections. Character drills and rounds under one second have no word speed. Legacy five-character-method values remain in the export, but not in the speed chart.
The optional pulse adapts to your recent error-free keystrokes: 10% extra time, 15% slower after errors, 40–150 beats per minute. It stops after three seconds without input and stays off during tests. You do not have to match it.
The hand guide, difficulty and pulse affect comparisons. Training patterns are descriptive, not a diagnosis or screening for ADHD, dyslexia or other conditions.
Your data & privacy
Without an account, your progress stays in this browser. With an account, we store your email address, settings and training results so you can continue on multiple devices. We only import your existing progress with your consent. An essential cookie keeps you signed in.
Exercises you have already opened also work without internet. With an account, new results are uploaded when you reconnect. Do not clear browser data before then, or exercises not yet uploaded may be lost.
Download a copy of your data under “History”. Delete your account and its stored progress under “Account”. Remove exercises saved without an account by clearing this website's data in your browser settings.
Selecting Lexend or Atkinson uses Google Fonts. Depending on your device, read-aloud may use an online service. Neither feature is required for typing practice.
Research & sources
The design is inspired by research. Neither typelapp itself nor its specific sequence has been tested in an effectiveness study. A different, consistent finger assignment is not automatically worse.
- Feit et al. (2016): consistent key-to-finger assignment and typing performance; not evidence for a retraining method.
- Shea & Morgan (1979): switching motor tasks; not a direct test of our typing training.
- Baddeley & Longman (1978): distributed typewriter practice; not an optimal number of minutes for typelapp.
- Bönstrup et al. (2019), Das et al. (2025) and Gupta & Rickard (2026): short-term gains during breaks are not conclusive evidence of lasting learning.
- McKay et al. (2022): no reliable general advantage of less frequent feedback.
- Guadagnoli & Lee (2004): a theoretical framework for difficulty, not validation of this learning path.
- Birkenbihl's ABC list: inspiration for personal memory cues, not evidence of effectiveness for finger paths.
- Yamaguchi et al. (2013): speed and accuracy; our pulse rule remains an unvalidated design idea.
- NICHD: Diagnosis: if difficulties persist in everyday life, professional assessment may help.