Watch closely…
Memory Test
A memory test shows you something briefly, then asks you to repeat it. Try visual, number or sequence memory, each one level harder than the last.
Watch the lit squares, then tap them.
A browser game for fun, not a medical, driving or safety test. It does not rate how smart you are. Your results stay on your device unless you submit a score to the Hall of Fame.
Your score is the highest level you cleared.
Next testChimp test
What 7 levels means
You cleared level 7 of the visual memory test under our rules. That depends on attention, practice and your screen size, and it says nothing about how smart you are. Compare it with your own earlier runs.
Hall of Fame
Visual memory test · highest level, higher is better
Claim your spot
Loading…
Best scores on this site, from browsers. Each run is checked against our rules for a human run; scores still depend on the device. How the Hall of Fame works · See all leaderboards
Memory test results: what your score means
Each of the three tests has its own score, and they are kept apart. In the visual memory test your score is the highest level you cleared. In the number memory test it is the number of digits in the longest number you typed correctly. In the sequence memory test it is the length of the longest sequence you repeated. If you cleared nothing, the score is 0.
The levels, the times and the lives are our own rules, chosen to make a short game. They are not taken from a published test, and the page does not say what a typical score is or rank you, because we have no source for one. A score moves with attention, sleep, noise, practice and screen size, so the useful comparison is your own earlier runs. It is a game for fun, not a measure of how smart you are.
For a different kind of speed, try the reaction time test or the chimp test, which is a number memory game on a grid.
Why your memory test score drops
- Tired or distracted. Memory games show it first. A run after a break is not comparable with one at the end of a long day, and a notification during the show phase can cost the whole level, because the time does not pause.
- Small phone grid. In the visual memory test the grid reaches 6 by 6 at level 10, and on a narrow screen a thumb can land on the neighbor. Hold the phone steady and tap with one finger.
- Typing slips. In the number memory test a swapped or missed digit counts as wrong. Read your answer once before you press Check.
- Double taps. A second tap on a square you already found is ignored in the visual test, but in the sequence test a second tap is the next step. Tap once and lift.
- Rushing. Nothing times your answer. Take a breath after the show phase ends and then start tapping or typing.
How the memory test works: visual, number and sequence
This memory test online runs in your browser, with no account and nothing uploaded. Pick a mode in the box above and press Start. Switch modes between runs; each mode keeps its own history on this device. All timings and rules below are our own choices.
Visual memory test
A square grid lights several squares at once for a moment, then they go dark. Tap every square that was lit, in any order. Level 1 lights 3 squares on a 3 by 3 grid, and each level lights one more. The grid grows to 4 by 4 at level 3, 5 by 5 at level 6 and 6 by 6 at level 10, and the squares stay lit for longer as there are more of them. A square that was not lit costs one of your 3 lives, and the same level comes back with a new pattern. The third lost life ends the run.
Number memory test
A number of 3 digits shows, then hides. Type it and press Check, or Enter on a keyboard. On a phone the number keypad opens. Each right answer makes the next number one digit longer and keeps it on screen for longer: 1 second plus 0.45 second for each digit, so about 2.4 seconds for 3 digits and 5 seconds for 9. There is no timer on typing. One wrong answer ends the run, and the page then shows the number it was. Spaces you type are ignored, and an answer with letters or nothing in it is not counted until you fix it.
Sequence memory test
A 3 by 3 grid lights one square at a time. When it stops, tap the squares in the same order. Each level plays the whole sequence again with one new step at the end, starting from one step. A step is lit for half a second with a quarter second dark after it, and the same square can come up twice in a row. One wrong tap ends the run.
| Test | Level 1 | Each level | The run ends |
|---|---|---|---|
| Visual memory test | 3 squares on a 3 by 3 grid | One more square; bigger grid at levels 3, 6 and 10 | On the third lost life |
| Number memory test | 3 digits | One more digit and 0.45 second more to look | On one wrong answer |
| Sequence memory test | 1 step | One more step at the end | On one wrong tap |
Nothing on this page flashes more than 3 times a second. The sequence test lights a step at most once every 0.75 second, or every 1.1 seconds if your device asks for reduced motion. Squares switch on and off at once; nothing fades or moves.
Read how this test measures for the sources of error and the rules behind each round.
Memory test FAQ
What is a memory test?
A memory test shows you something to remember, hides it, and checks what you can bring back. The three tests on this page measure short-term recall: where squares lit up, which digits appeared, and in what order squares flashed. They are games for fun, not clinical tools, and a score says nothing about your health.
How does the visual memory test work?
A grid lights a few squares at once, then they go dark. Tap every square that was lit, in any order. Each level lights one more square, and the grid gets bigger at levels 3, 6 and 10. A square that was not lit costs one of your 3 lives, and the third lost life ends the run. Your score is the highest level you cleared.
How does the number memory test work?
A number shows for a few seconds, then hides. Type it and check it. A right answer makes the next number one digit longer and gives you a little longer to look. One wrong answer ends the run. Your score is the number of digits in the longest number you typed correctly.
How does the sequence memory test work?
A 3 by 3 grid lights one square at a time. When it stops, tap the squares in the same order. Each level plays the same sequence again with one new step at the end. One wrong tap ends the run, and your score is the length of the longest sequence you repeated.
Is there a verbal memory test here?
No. This page has three tests: visual, number and sequence memory. None of them uses words or spoken recall, so it says nothing about verbal memory.
What is a good memory test score?
This page has no sourced norm, so it gives no label, no average and no ranking. The levels, digits and steps are counts under our own rules, so they cannot be compared with other sites or with published studies. Compare a run with your own earlier runs, on the same device and at a similar time of day.
How can I do better on a memory test?
Group what you see instead of holding items one by one: squares as a shape, digits in chunks of two or three, sequence steps as a path across the grid. Say digits quietly to yourself. Play when you are rested, with notifications off, and take a short break after a bad run instead of retrying straight away.
Do memory games improve memory?
We cannot say. This page has no source showing that playing memory games for adults, or for anyone, carries over to daily life, so it makes no claim either way. What you can expect is practice at these games, such as learning to chunk digits, which can change your score here. That is a different thing from memory in general.
Where the memory test comes from
The three games on this page are our own and follow our own rules. They borrow ideas from two older kinds of memory task, and this section says what the sources we could read do and do not say about each.
Number memory test. It is a digit-span-style task: a string of digits is shown, hidden, and repeated back. In 1956 George Miller reviewed experiments of this kind and wrote that for many kinds of test material the span of immediate memory "is about seven items in length", and that the span is better counted in chunks than in single items.1 He kept the number loose, calling the repeated sevens "pernicious, Pythagorean coincidence".1 It was a summary of other people's results, not a norm, and we do not use it as one.
In 2001 Nelson Cowan described Miller's number as "a rough estimate and a rhetorical device" rather than a real capacity limit, and argued for about four chunks when rehearsal and chunking are blocked. He notes that others had suggested three to five chunks.2 We read only the abstract of that paper, so we say no more than it does. The estimates differ with the method, and a game that lets you rehearse and group digits, as ours does, is not set up to measure either one.
Sequence memory test and visual memory test. Both are related to the Corsi block-tapping task. In the traditional version, nine blocks sit on a board, the examiner taps a sequence starting with two blocks, and the person taps it back in the same order, or in reverse. When enough sequences of one length are repeated correctly, the next one is a block longer.3 Our sequence test uses the same idea of a growing order, on a 3 by 3 grid of squares. The visual test asks for lit squares in any order, so it is a looser relative.
The Corsi task has been widely used in clinical and experimental studies, and one review found that setups and scoring differed between labs.4 A standardized administration and scoring procedure for it has been published.5 Its published clinical figures are not quoted here, because our games are not that procedure.
What this means for your score. Our games use their own rules, laid out in the table above, so they are not those lab procedures and are not clinical tools. We found no published norm for web memory games like these, so this page gives none, and a run here cannot be compared with a study.
Sources
- Miller. The magical number seven, plus or minus two: some limits on our capacity for processing information. Psychological Review 63(2):81-97 (1956). https://psychclassics.yorku.ca/Miller/ (read: full text, checked Nov 8, 2026)
- Cowan. The magical number 4 in short-term memory: a reconsideration of mental storage capacity. Behavioral and Brain Sciences 24(1):87-114 (2001). doi:10.1017/S0140525X01003922. https://doi.org/10.1017/S0140525X01003922 (read: abstract only, checked Nov 8, 2026)
- Brunetti, Del Gatto, Delogu. eCorsi: implementation and testing of the Corsi block-tapping task for digital tablets. Frontiers in Psychology 5:939 (2014). doi:10.3389/fpsyg.2014.00939. https://doi.org/10.3389/fpsyg.2014.00939 (read: full text, checked Nov 8, 2026)
- Berch, Krikorian, Huha. The Corsi block-tapping task: methodological and theoretical considerations. Brain and Cognition 38(3):317-338 (1998). doi:10.1006/brcg.1998.1039. https://doi.org/10.1006/brcg.1998.1039 (read: abstract only, checked Nov 8, 2026)
- Kessels, van Zandvoort, Postma, Kappelle, de Haan. The Corsi Block-Tapping Task: standardization and normative data. Applied Neuropsychology 7(4):252-258 (2000). doi:10.1207/S15324826AN0704_8. https://doi.org/10.1207/S15324826AN0704_8 (read: abstract only, checked Nov 8, 2026)
Each source was read on Nov 8, 2026, in full or, where marked, in the abstract only; what we say about an abstract-only source goes no further than the abstract. Miller's paper has no DOI we could confirm, so its entry links to a reproduction of the paper on the York University Classics site, which is the copy we read.
Method v1.0 · Updated Nov 8, 2026