TYPING PERFORMANCE DIAGNOSTIC
Find out what's actually
slowing your typing down.
Most typing tests give you one number. This one measures you under three controlled conditions — plain words, real punctuation, and text full of numbers and symbols — and tells you which one is costing you speed.
- About 3 minutes
- No signup
- Results stay in your browser
Bottleneck found
Symbols and numbers are what slow you down.
56% slower on text with digits and brackets than on common words. Raw speed is not the limit here.
Measure, diagnose, practise, verify
Every step hands you the next one. You never have to work out what to practise or find your way back to the test.
- 01
Measure
Three fixed 45-second conditions. Each changes exactly one thing about the text, so the difference between them means something.
- 02
Diagnose
One sentence naming what costs you the most speed, followed by the measurements it came from and the thresholds behind it.
- 03
Practise
A drill built from that result, not a generic word list. The page tells you which finding it targets and why.
- 04
Verify
Re-run the same protocol and see the before and after — or be told plainly that there is not yet enough evidence to say.
What it measures — and when it refuses to
Every figure has a formula and a minimum amount of evidence. Below that threshold the report says so instead of showing a number, because a confident-looking statistic drawn from four keystrokes is worse than no statistic at all.
- Speed per condition
- Net WPM on plain words, on natural prose, and on text with digits and symbols.
- Reported whenever that stage completes cleanly.
- Punctuation penalty
- How much slower natural prose is than plain words, as a share of your own baseline.
- Named as a bottleneck from 15% downward.
- Complexity penalty
- The same comparison against numbers, currency, brackets and identifiers.
- Named as a bottleneck from 15% downward.
- Correction cost
- Time spent noticing a mistake, deleting back to it and retyping, as a share of the stage.
- Withheld below 3 completed corrections.
- Attempt accuracy
- Correct keystrokes as a share of every keystroke, including the ones you fixed.
- Flagged below 95%.
- Pace variation
- How evenly your output arrives across five-second windows of the flow stage.
- Withheld below 40 first attempts.
- Repeated mistakes
- Which character was expected and which one arrived, pooled across all three conditions.
- Withheld below 10 chances and 3 errors.
Free practice
Choose a language for a free typing exercise
These practice pools use different text and are not part of the standard English benchmark.
YOUR RESULTS
Your history stays in this browser
Nothing is uploaded and there is no account. Results are saved locally so you can compare a later run against an earlier one, and you can clear them at any time.
What this does not claim
A diagnostic is only worth anything if it is honest about its own limits. These are ours, stated before you spend three minutes on it.
No population percentile
There is no verified dataset of other people’s results, so no rank, percentile or "faster than X%" is shown anywhere. Comparisons are against your own earlier runs only.
Two runs are not a trend
A before and after states a difference between two measurements. It does not claim improvement, because typing varies between sessions and two points cannot separate the two.
English, physical keyboard
The diagnostic protocol covers English on a physical keyboard. Other languages and touch devices are available as practice, and are not scored against this standard.
Local-first history
Keystroke traces are scored in your browser and never uploaded. Only bounded summaries are stored, in this browser, and you can clear them from the practice page.