HumanSpec Benchmarks
A score is more useful with context.
Use these reference points to understand what a result means, then compare your own range under consistent conditions. They are orientation points—not a live leaderboard and not a judgment about ability.
Reaction Time Test
Measure how quickly you respond to a visual signal across five original rounds.
Typical reference point
250 ms
Reaction time
Click Speed Test
See how many accurate clicks you can make in five seconds.
Typical reference point
6.2 CPS
Clicks per second
Typing Speed Test
Test your words per minute and accuracy with a short timed passage.
Typical reference point
42 WPM
Words per minute
Number Memory Test
How many digits can you hold in your working memory?
Typical reference point
7 digits
Digits recalled
Visual Memory Test
Remember the highlighted tiles and reproduce the pattern as it grows.
Typical reference point
Level 4
Pattern level
Sequence Memory Test
Remember a short sequence of digits, then reproduce it after it disappears.
Typical reference point
3 digits
Sequence recalled
Stroop Test
Practice selective attention by responding to a color while ignoring a conflicting word.
Typical reference point
Practice task
Color-word accuracy
N-Back Test
Try a short working-memory updating task with a simple one-back sequence.
Typical reference point
Practice task
Match accuracy
Aim Trainer
Click changing targets to practice visual-motor accuracy in a short browser task.
Typical reference point
5 targets
Targets hit
Reaction time by age
What the evidence can—and cannot—tell you
Searches for average reaction time by age often produce one neat number for each decade. That presentation is convenient, but it hides important differences in task design. A simple visual reaction task, a choice reaction task, and a browser test do not measure exactly the same signal. Display refresh, input latency, browser scheduling, anticipation, sleep, attention, and practice all move the observed result.
Research on adult reaction time does support a broad pattern: average responses tend to become slower with age, while individual distributions overlap substantially. The pattern is useful for understanding population-level variation, not for assigning a score to a person. HumanSpec therefore does not publish invented age norms or imply that a single result can diagnose ability, health, or intelligence.
| Age band | Published evidence | HumanSpec sample | How to use it |
|---|---|---|---|
| Under 20 | Protocol-dependent; use the cited study rather than a single universal norm. | Not published yet | Compare repeated sessions on the same setup. |
| 20–29 | Protocol-dependent; young-adult samples are commonly used as a reference group. | Not published yet | A faster result is not automatically a better overall performance. |
| 30–39 | Reaction time distributions vary with task choice, device, and sample composition. | Not published yet | Use your own median and range as the primary baseline. |
| 40–49 | Age-related trends are visible in adult survey data, but they are not individual cutoffs. | Not published yet | Keep display and input latency stable before comparing sessions. |
| 50+ | Published adult data generally show wider and slower distributions with increasing age. | Not published yet | Context matters more than a single pass/fail number. |
The table is an interpretation framework, not a normative dataset. HumanSpec will only publish age-stratified means and standard deviations after collecting and documenting a consented sample with a pre-declared protocol.
What is a good score?
Use percentiles within your own sessions
A “good” score is relative to the task and the conditions. For reaction time, a smaller number can mean a faster response, but an unusually small number can also reflect anticipation or a lucky round. For typing and click speed, accuracy matters alongside speed. For memory tasks, a familiar pattern or practice effect can change the result.
For a practical baseline, take three to five attempts under the same conditions and record the median, the middle half of your results, and the date. You can then describe a session as below, near, or above your own usual range. This is more honest than calling one result “excellent” without knowing the test protocol or the person’s setup.
Use the individual test pages for the current orientation ranges, then read the science and limitations before comparing results across devices. The personal baseline guide explains a simple log you can reuse.
How to compare fairly
A repeatable routine makes a personal baseline more useful than a leaderboard-style claim.
Keep the setup stable. Use the same browser, display, keyboard, mouse, and connection whenever possible.
Look for direction. Improvement means becoming more consistent, not simply producing the smallest number.
Complete several attempts. A small sample is more useful than one unusually fast or slow round.
Look for direction. Improvement means becoming more consistent, not simply producing the smallest number.
Method and limitations
Why browser benchmarks need context
HumanSpec runs short tasks in an ordinary browser. That makes the tests convenient and repeatable, but it also means the result includes more than the participant’s response. The screen may take time to refresh, the operating system may schedule the tab between frames, and a mouse, keyboard, touchpad, or touchscreen may report input differently. Network speed is usually less important than local rendering and input timing, but browser load and background work can still affect a session.
Each score on this page is a reference point for a specific task, not a claim about the general population. Keep the browser, display, input method, seating, and instructions stable. Complete several rounds, record the median or typical range, and compare your own sessions over time. If you change hardware, treat the first session as a new baseline rather than mixing the numbers.
These tests are educational and recreational. They do not diagnose a condition, measure IQ, establish fitness for work or school, or predict athletic performance. If you have a health concern, use an appropriately validated assessment with a qualified professional.
Cross-test reference points
One framework for several signals
Reaction time, click speed, typing speed, number memory, and visual memory answer different questions. A quick click score is not a substitute for a reaction-time score, and a memory span is not a general measure of intelligence. The useful comparison is within the same task: keep the setup stable, repeat the session, and watch whether your typical range changes.
Use the cards above as entry points to the individual test pages. Each page explains the metric, shows an orientation range, describes how the score is produced, and links to related practice and FAQ content. This creates a clearer path from an information search to a test session without presenting a thin click-bait page as a scientific assessment.
Reaction Time Test
Reaction time: 250 ms. Read the method and try a repeatable session.
Click Speed Test
Clicks per second: 6.2 CPS. Read the method and try a repeatable session.
Typing Speed Test
Words per minute: 42 WPM. Read the method and try a repeatable session.
Number Memory Test
Digits recalled: 7 digits. Read the method and try a repeatable session.
Visual Memory Test
Pattern level: Level 4. Read the method and try a repeatable session.
Data status
Transparent about what is measured
The values shown on the test cards are orientation points for the current HumanSpec tasks. They are not calculated from a hidden population database, and HumanSpec does not currently claim to have a representative age-stratified sample. That distinction matters: a number can look precise while still being a poor estimate of a population average if the task, device, or sample is not documented.
If HumanSpec publishes an aggregate dataset in the future, the protocol should be declared before collection. It should record the task version, browser, operating system, display refresh rate, input method, age band, number of attempts, and whether the session was completed on a desktop or touch device. The report should include sample size, mean, standard deviation, median, interquartile range, missing data, and exclusion rules for each group.
Until then, use this page for definitions and interpretation, and use your own repeated sessions for a personal baseline. That is a smaller claim, but it is a more reproducible one.
Important limitation
HumanSpec benchmarks are educational reference points collected for a browser experience. They are not population norms, medical cutoffs, IQ scores, or employment and academic assessments. Sleep, practice, attention, display latency, and input hardware can all change a result.
Read the science and limitationsBenchmark FAQ
Are HumanSpec benchmarks medical or IQ scores?
No. HumanSpec benchmarks are educational browser reference points, not medical, psychological, IQ, employment, or academic assessments.
What is a good reaction time?
A good result depends on the task, device, display, input method, and practice. Use several attempts to establish your own repeatable range instead of treating one number as a universal cutoff.
Does reaction time get slower with age?
Large adult studies generally report slower average reaction times with increasing age, but the size of the difference depends on the protocol and the sample. It should not be used to judge an individual from one browser session.
How can I compare my result fairly?
Use the same browser, display, keyboard, mouse, and connection where possible, then compare several attempts instead of one outlier.
Why can my result change between sessions?
Sleep, attention, practice, browser load, display latency, and input hardware can all change a browser test result.
Are the age bands HumanSpec population norms?
No. HumanSpec has not published an age-stratified normative dataset. The age bands on this page are a reading framework, while the cited research explains broader population patterns.
References
Research behind the context
- Woods, D. L., Wyma, J. M., Yund, E. W., Herron, T. J., & Reed, B. (2015). Factors influencing the latency of simple reaction time. Frontiers in Human Neuroscience, 9, 131. doi.org/10.3389/fnhum.2015.00131
- Der, G., & Deary, I. J. (2006). Age and sex differences in reaction time in adulthood: Results from the United Kingdom Health and Lifestyle Survey. Psychology and Aging, 21(1), 62–73. doi.org/10.1037/0882-7974.21.1.62
- Welford, A. T. (1980). Choice reaction time: Basic concepts. Academic Press. This foundational reference explains why task design and response selection affect reaction-time interpretation.
References describe general measurement and adult reaction-time patterns. They are not a validation study of HumanSpec, and HumanSpec does not claim that these sources establish clinical or population norms for this website.