Brain Test

Measure it, then watch it change

Short tests that run straight in your browser with nothing to install. Each one takes a minute or two, your result is saved so you can compare it with your own earlier attempts — and nothing here claims to measure your intelligence.

🧠

The full brain test — all six axes

Six tasks, one from each axis, in about eight minutes. You finish with a six-sided profile, your raw numbers beside it, and a measurement the single tests cannot give you: delayed recall — what survives in memory eight minutes after you stopped looking.

Six axes Profile chart Delayed recall ≈ 8 min
Start the session →
⚡

Reaction time

A colour is named, then the squares appear at once — find it and tap it. Starts with two squares and ends with sixteen, so it measures decisions, not just fingers.

SpeedDecision cost≈ 2 min
🧩

Visual memory

Squares light up in order and you repeat the order. Each success makes the sequence longer, until two misses end it.

Working memorySequence≈ 1 min
🧠

Matrix reasoning

A 3×3 grid with one cell missing and six near-identical options. Find the rule, then the missing piece. Generated, never hand-drawn — and not an IQ test.

LogicFluid reasoning≈ 3 min
🎯

Stroop test

BLUE printed in green ink: tap the ink colour, not the word. We report what the conflict costs you, not your raw speed.

FocusInterference≈ 2 min
🔄

Task switching

One digit, two rules that keep swapping: odd or even, bigger or smaller than five. The number you get is what changing the rule costs you.

FlexibilitySwitch cost≈ 2 min
🔃

Mental rotation

Two shapes: is the second the first rotated, or its mirror image? Besides accuracy we measure what every 45 degrees of rotation costs you.

SpatialRotation cost≈ 2 min

What these tests are — and what they are not

They are browser adaptations of tasks used in research, built to be quick, honest and free. Each test page says where its idea came from and what it cannot tell you.

They are not IQ tests, not clinical screening, and not a diagnosis of anything. You will find no percentile here either: we have no population sample, so you beat 87% of people would be invented. What you get instead is the comparison that is actually sound — you against your own earlier attempts on the same device.

The reasoning behind every choice, the exact bounds of the chart scale, and a dated log of every correction we have made are all in the methodology.