Pick a duration and type as accurately as you can. Timer starts on your first keystroke. Track your net WPM and accuracy in real time.
Select a duration, then type. The clock starts with your first keystroke and ends automatically. Paste is disabled.
WPM uses the standard five-character word. Net WPM counts correct characters only; final accuracy measures errors still present at the end. Mistyped keys include errors corrected later.
Actual typing samples, not a decorative waveform.
A typing speed test measures how quickly and accurately you reproduce written text using a keyboard. ReflexBenchmark's free online typing test gives you a passage to copy and reports your speed in words per minute (WPM). It also separates the number of characters typed from the characters matching the passage, so you can see the difference between moving quickly and typing accurately.
Unlike a task that asks you to invent text, this typing practice test uses a supplied passage. Your job is to type its characters in order, including spaces and punctuation. The browser highlights characters as you go, shows a running WPM estimate and provides a final result with error information. Use it to monitor your keyboard practice; it is not a standardized certification.
The passage extends throughout the timed run. Paste and drop are disabled; switching away from the browser tab interrupts the run and prevents a score submission for that attempt. The New passage and Try Again controls let you reset and start another attempt.
For this test, one standardized word equals five characters, not one dictionary word. The browser counts characters matching the reference passage at their positions. Gross WPM uses the total number of characters still entered at the end of the run; net WPM uses the number of those characters that match. Each result is divided by five and scaled from the test duration to a one-minute rate. Both displayed WPM values are rounded to whole numbers.
Final accuracy is the percentage of entered characters that match the passage: correct characters ÷ entered characters × 100, rounded to a whole percent. The initial 100% display before typing is not a measured accuracy result. Other websites may calculate WPM differently.
Suppose you finish with 240 entered characters, of which 228 match the displayed passage. Your gross speed is 240 ÷ 5 = 48 WPM. Your net speed is 228 ÷ 5 = 45.6, displayed as 46 WPM. Your final accuracy is 228 ÷ 240 × 100 = 95%.
For a 30-second run, the same number of correct characters would produce a different WPM because the result is scaled to 60 seconds. These numbers are an illustrative calculation, not published averages for website visitors.
Net WPM is the primary score shown in large type and sent to the website's score-saving API. Gross WPM reports all characters remaining in the input, whether correct or incorrect. The gap helps identify when fast typing also left unresolved errors.
The result also distinguishes mistyped keystrokes, Backspace actions and errors remaining. Mistyped keystrokes include incorrect characters entered and later corrected. Errors remaining count characters still mismatched at the end of the test. Backspace actions count character deletions detected as you edited. Do not add these counters together as unique mistakes.
The measured speed-over-time chart uses samples collected during the actual run. Longer modes normally sample at about five-second intervals; the 15-second mode samples more often. It shows changes in your running net WPM, not a laboratory-quality performance trace.
Different durations answer different practice questions. A 15-second typing test is convenient for a quick warm-up but may be sensitive to the opening words and your initial pace. A 30-second typing test provides a short challenge with more text than the briefest mode. The default 60-second typing test gives you a full minute to balance speed, accuracy and error correction. A 120-second typing test offers a longer interval in which consistency and concentration become more noticeable.
Do not compare all four scores as if they came from the same test conditions. Short and long runs can feel different, and the website separates comparable rankings by selected duration. Track improvement using a consistent duration and setup.
There is no single WPM value that describes a good result for every person, job, language or keyboard. A useful score depends on what you want to do. For a beginner, fewer errors and a comfortable typing rhythm may matter more than reaching a particular number. Someone entering long passages may care about how stable their net speed remains, while a person practicing short keyboard tasks may prefer a shorter session.
This test does not establish a verified population-wide average or expert threshold. When the site has enough comparable saved attempts in the same timed mode, it may show an estimated percentile among those recorded website attempts. Those submissions are not a representative global sample. If ranking is pending, your measured WPM and accuracy remain valid as local test results without a population comparison.
Touch typing can be a useful goal, but layouts and accessibility needs vary. No improvement percentage is guaranteed.
Keyboard and device: Keyboard layouts and input methods vary in feel and familiarity. A score taken on a phone is not necessarily comparable with one taken on a desktop keyboard.
Passage and punctuation: Text with unfamiliar words, capitalization or punctuation may slow you down. This site rotates through practice passages and expands the text for longer runs. Passages may differ between attempts even within the same mode.
Timing and attention: The browser measures elapsed time using its own clock. Background-tab interruptions end a run rather than awarding an unreliable score. Distractions and corrections can affect the result.
ReflexBenchmark provides an informal browser-based typing benchmark, not an official hiring, educational or medical assessment. The reported WPM reflects the website's character-matching formula and selected duration. It is not a guarantee that you will write, transcribe or code at the same speed in an unrelated task. A recorded score can also be affected by browser behavior and input methods.
Saving a score depends on an API response. If saving fails, the browser may still display a calculated personal result, but that alone does not confirm database storage. A comparable rank needs enough recorded attempts from the matching time mode; percentile availability should never be mistaken for a verified worldwide ranking.
Try the 10-Key Typing Test to practice numerical entry on a keypad, or the Processing Speed Test for a different type of timed matching. The Reaction Time Test measures responses to a visual cue rather than sustained text entry, while the Attention & Focus Test focuses on another aspect of responding accurately under task demands. Each test has its own scoring method, so compare your progress within the same tool.
It counts five characters as one standardized word. Net WPM uses correctly matched characters, divides by five and scales the result to one minute using your chosen duration.
Net WPM uses characters that match the passage; gross WPM uses all entered characters remaining at the end. A difference between them reflects unresolved typing errors.
The timer starts when you enter your first character, not when the page loads. The run automatically ends after 15, 30, 60 or 120 seconds, depending on your selection.
Final accuracy is the number of correctly matched characters divided by the number entered, multiplied by 100 and rounded. Mistakes corrected before the end are counted separately.
Choose one duration and use it consistently. The 60-second mode is the default, while 15 and 30 seconds provide shorter practice and 120 seconds tests a longer stretch.
The site needs enough comparable saved attempts in your selected time mode before showing a rank. Pending does not mean the WPM calculation failed or that your score was poor.
Paste and drop are disabled. If you switch away from the page while a run is active, the attempt is interrupted and no score is submitted for that run.