Free cognitive tests ๐งฉ
No signup, no install, nothing sent to a server. Just open one and start.
Vision & Colour Check
Landolt-C acuity, then Ishihara-style plates where only hue reveals the digit
Open ๐Full Cognitive Test
Memory, focus, speed, logic and spatial ability in five minutes โ as a radar chart
Open โกReaction Time
Five rounds, average in milliseconds, plus how much of it is your hardware
Open ๐งN-back
1, 2 and 3-back working memory, scored so false alarms actually cost you
Open ๐จStroop Test
Name the ink colour, ignore the word. 30 seconds of inhibitory control
Open โจCorsi Block Test
Watch the sequence, tap it back. The standard measure of visuospatial memory span
Open ๐Mental Rotation
Rotation or mirror image? Every pair is guaranteed chiral, so silhouette will not save you
Open ๐Memory Match
Concentration, scored for wasted flips as well as speed
Open ๐ขQuick Math
Timed arithmetic where wrong answers subtract, so guessing costs you
Open ๐จColour Spotting
One tile differs, and the difference shrinks every level
Open ๐Number Merge
2048-style planning under forced randomness, with the corner strategy explained
Open ๐งญBig Five Personality Test
The five-factor model researchers actually use, with reverse-scored items so agreeing with everything does not inflate your profile
Open ๐งฉCareer Interest Test (RIASEC)
Holland's six vocational types and your three-letter code. Interest is not ability, and we say so
Open โ๏ธBall Weight Ordering
Sort identical-looking balls by weight, with a weighing budget set at the theoretical minimum
Open ๐ผTower of Hanoi
Scored against the 2โฟโ1 minimum, with a mechanical rule that solves it without planning
Open ๐Task Switching
Position decides which rule applies. Measures the switch cost that makes multitasking expensive
Open ๐Quantity Comparison
Which side has more? Dot sizes are jittered so judging by area will not work
Open โRock Paper Scissors Reflex
Win, lose or draw on command. Losing on purpose is the hard part
Open ๐Balloon Risk Task (BART)
Pump for points, bank before it bursts. A standard behavioural measure of risk-taking
Open ๐Iowa Gambling Task
Four decks, 24 draws. You will start avoiding the bad ones before you can say why
Open ๐งชPotion Brewing
Work out the hidden rule from red/blue feedback. Testing to confirm gets you nowhere
Open ๐ขNumber Pressing
Tap 1-9 in the order the rule demands โ and the rule keeps changing
Open ๐คFinding Common Ground
Three people, five seconds, one thing they all agree on
Open ๐ค๏ธPipe Path Puzzle
Rotate tiles to connect both sides. Propagate the forced tiles before you guess
Open ๐จWhack-a-Mole
Hit moles, avoid bombs. A go/no-go inhibition task in an arcade costume
Open ๐Snake
The difficulty is space, not speed โ the maze is one you built yourself
Open ๐งฑBrick Breaker
Where the ball hits the paddle sets its angle. That is the entire game
OpenReading
Eight cognitive tests, explained
What Stroop, N-back, Corsi, BART and four more actually measure, and where each came from
Guide ๐Does brain training actually work?
Near vs far transfer, what happened with N-back, the 11,000-person trial, and what does help
Guide ๐งHow to improve your memory
Five techniques ordered by evidence โ and why highlighting does almost nothing
Guide ๐ฏHow to improve your focus
Sleep, multitasking and the phone-in-the-room effect, separated from the advice that just sounds right
GuideWhat this is
A small collection of classic cognitive tasks โ the same paradigms used in psychology labs โ rebuilt to run in a browser without an account. Reaction time, N-back, Stroop, a spatial span task, mental rotation, and an Ishihara-style colour check.
Everything is free. There is no signup, no install, and your results are stored in your browser only (localStorage) โ nothing is sent to a server unless you choose to submit a score to a leaderboard. The site is funded by ads.
What it is not
Worth being direct about this, because a lot of sites in this space are not.
- It is not an IQ test and does not produce one. Five short browser tasks cannot substitute for a standardised battery administered under controlled conditions.
- It cannot diagnose anything. Not attention disorders, not colour blindness, not visual acuity. If you have a concern, see a professional.
- It will not make you smarter. The evidence for far transfer from brain training is much weaker than the marketing suggests โ we say so plainly on the N-back page. What you reliably gain is familiarity with the task format, which is genuinely worth something when you meet these tasks in a test.
What it is good for: seeing your own relative strengths, and watching how sleep and fatigue move your own numbers. Those effects are large and easy to observe.
A note on measurement
Browser-based measurement has real limits and we would rather name them than paper over them.
- Reaction time includes your hardware. A 60Hz display adds up to 16ms before the stimulus is even visible, and touch panels add more. Compare your own device to itself, not to a stranger.
- Vision cannot be scored as 20/20. Screen size and viewing distance are unknown to the page, so the check reports stages cleared rather than inventing a figure.
- No percentiles yet. Ranking you against others requires a large sample; with a small one the number swings every time someone new arrives. When there is enough data, we will turn it on.
About the rest of the site
CogPop began as a Korean-language site. All 27 games and tests are now fully available in English โ the pages above. What remains Korean-only is the job-application material: an aptitude mock exam and a typing test whose questions are Korean text, and a set of guides written for the Korean graduate hiring market. The four cognitive-science guides above are in English. None of it would be useful translated.
The Korean site has those as well, if you read Korean.