A Human Benchmark alternative with 20 tests
Human Benchmark is where most people first measure their reaction time, and it deserves the credit for that. If you’ve worked through its tests and want a wider set — or rankings computed from a pool that actively filters implausible scores — here is a factual side-by-side of what each site does.
Human Benchmark details reflect their public site as of August 2026. Check theirs for the current picture — we’ll correct anything that drifts.
What’s actually different
Rankings computed from a defended pool
A public leaderboard is only as meaningful as the scores on it, and any leaderboard that accepts whatever a browser sends attracts exactly the entries you’d expect. Communities around benchmark sites have reported injected scores for years, and the same would happen to us if we ran one undefended. MindBenchmark doesn’t publish a top-scores table at all. Instead, every submission goes through a server-issued single-use session, gets checked against per-test plausibility bounds, is rate-limited per device and per network, and carries a trust score. Your percentile is computed against what survives that.
More of the cognitive map
Reaction time and memory span are the easy ones to build, so most sites stop there. MindBenchmark covers 20 tests including interference control (Stroop), sustained attention, task switching, mental arithmetic, and mental rotation — the abilities where your speed score tells you nothing.
A number, then an explanation
Every result comes with a tier label, a percentile, and — for common scores — a written page explaining what that band means and what moves it. These are measurements of performance on a specific task, not assessments of intelligence or health, and we say so on every page.
A reason to come back tomorrow
The daily challenge picks one test per day — the same one for everyone — and tracks your streak, which is a better way to see real change than replaying your favorite test until you get a lucky run.
Where Human Benchmark still wins
It has been running far longer, it’s the name people recognize, and it has collected a far larger volume of raw attempts than we have. If you want the number everyone else quotes, or you like a visible leaderboard to chase, that’s a fair reason to stay. Our pool is smaller and younger — we’d rather tell you that than pretend otherwise.
The same tests, here
Parameters differ between platforms, so treat these as equivalents rather than identical tests.
Reaction Time
Reaction Time Test
Sequence Memory
Sequence Memory Test
Aim Trainer
Aim Trainer
Number Memory
Number Memory Test
Verbal Memory
Verbal Memory Test
Chimp Test
Chimp Test
Visual Memory
Visual Memory Test
Typing
Typing Speed Test
And 12 they don’t have
Stroop Test
focusAttention Span Test
focusTask Switching Test
focusMental Math Test
numeracySpatial Reasoning Test
numeracyPattern Recall Test
numeracyPaired Memory Test
memoryPeripheral Vision Test
focusColor Blind Test
focusNumber Reaction Test
speedVisual Reaction Time Test
speedClick Speed Test (CPS)
speedFrequently asked questions
Is MindBenchmark free?
Yes. All 20 tests are free to take, with no sign-up, no trial, and no ads on any test page.
Do I need an account to save my scores?
No. Your personal best for each test is stored in your own browser. There is no account system to sign up for.
How are the percentiles calculated?
Each score is compared against the pool of scores other people have submitted, refreshed nightly. Submissions pass server-issued session checks, plausibility bounds, and rate limits before they count, and every row carries a trust score so implausible entries can be held back from the rankings.
Can I still compare my scores if I used another site first?
Roughly, but not exactly. Test parameters differ between sites — round counts, timing rules, and what counts as an error — so a score from one platform is not directly interchangeable with the same-named test elsewhere. Take a couple of runs here to establish your own baseline.
Does MindBenchmark work on a phone?
Yes, though the timing-sensitive tests are fairer on a desktop with a mouse. Typing and aim tests in particular are built for a keyboard and pointer.
“Human Benchmark” is a trademark of its respective owner and is used here only to identify that site for comparison. MindBenchmark is independent, and is not affiliated with, endorsed by, or sponsored by Human Benchmark.