Enter a score and every number updates at once — percentile, band, rarity and the range it realistically sits in.
Wechsler (WAIS/WISC) and Stanford-Binet 5 both use SD 15.
Hover the chart to read any point.
| IQ | Percentile | Rarity | Band |
|---|---|---|---|
| 70 | 2.275% | 1 in 44 below | Below average |
| 85 | 15.866% | 1 in 6.3 below | Average |
| 100 | 50.000% | 1 in 2 | Average |
| 115 | 84.134% | 1 in 6.3 above | Above average |
| 130 | 97.725% | 1 in 44 above | Gifted |
| 145 | 99.865% | 1 in 741 above | Highly gifted |
Every test carries measurement error. Drag reliability to see the range widen.
A score of 100 from a test with reliability .90 really means 91–109 — which stays inside Average.
Every tool here needs one thing: a number worth analysing. Converters, percentiles, comparisons and curves are all arithmetic performed on a score you already hold. They are exact, and they are only ever as good as the measurement you feed them.
Every calculator here runs off the same constants and the same curve, so no two can ever contradict each other. Start with the question you actually came with.
| Tool | Answers | You enter | You get |
|---|---|---|---|
| Percentile Calculator | What does my score mean? | A score or a percentile | Percentile, rarity, band, 95% range |
| Bell Curve | Where do I sit among everyone? | A score, dragged on the curve | Position, bands, area between two points |
| Score Converter | Why is my number different elsewhere? | A score and its scale | The same score on SD 15, 16 and 24 |
| IQ Comparison | Is the gap between two scores real? | Two scores and a reliability | A verdict, plus the share of people between |
| Old Test, New Norms | Is my old score still valid? | A score and its norming year | Today’s equivalent, with an honest range |
| IQ and Age | Does IQ change as I get older? | An age and a score | Reported IQ vs raw reasoning and knowledge |
| Regression to the Mean | Will my child score like me? | Two parent scores | The child’s likely range, and why |
Most IQ calculators give you a bare number. A bare number cannot be interpreted, and it is the reason two sites can hand the same person two different answers.
The same ability is 130 on a Wechsler test, 132 on an older Stanford-Binet, and 148 on Cattell III B. An IQ number without its standard deviation is not interpretable.
No test returns a single true value. A score of 120 from a test with reliability .90 really means 111–129. Reporting the point and hiding the range is the most common omission in this category.
Above three standard deviations the curve is an extrapolation, not a count. We cap what we report there rather than printing a false precision like “1 in 31,574”.
They are usually not disagreeing about your ability. They are printing it on different rulers and not saying which.
| Wechsler · SD 15 | Stanford-Binet L-M · SD 16 | Cattell III B · SD 24 | Percentile |
|---|---|---|---|
| 85 | 84 | 76 | 15.866% |
| 100 | 100 | 100 | 50.000% |
| 115 | 116 | 124 | 84.134% |
| 130 | 132 | 148 | 97.725% |
| 145 | 148 | 172 | 99.865% |
Mensa’s entry requirement is the 98th percentile — not a score. That is why the qualifying number is 132 on a Wechsler test and 148 on Cattell III B. Three numbers, one ability.
Note that Stanford-Binet 5 — the current edition — uses SD 15, not 16. SD 16 belongs to the retired L-M and fourth editions. Sites that apply SD 16 to a modern Stanford-Binet score report the wrong percentile.
Everything on this page is downstream of a single measurement. A percentile computed from a five-minute quiz is calculated just as precisely as one from a properly normed assessment, prints just as confidently, and means nothing at all. The tools cannot check their own input — that part is up to you.
Certified by the International Intelligence Federation — scored on the standard scale, mean 100 and standard deviation 15.
Your percentile is the share of the population that scores below you. An IQ of 115 sits at the 84.1st percentile, which means roughly 84 people in 100 score lower and about 1 in 6.3 score that high or higher. Percentile is more useful than the raw IQ number because it does not depend on which scale the test used. The IQ percentile calculator gives it exactly for any score.
Because they use different standard deviations. The same ability is 130 on a Wechsler test, 132 on an older Stanford-Binet, and 148 on Cattell III B. All three are the 97.7th percentile. An IQ number is only meaningful when the scale it was measured on is stated alongside it — the score converter shows any score on all three at once.
Less precise than most people assume. Every test has a standard error of measurement, so a result is a range rather than a point. A score of 120 from a test with a reliability of .90 corresponds to a 95% confidence range of about 111 to 129. This means a five-point difference between two people, or between two sittings, is not a meaningful difference. The comparison tool checks whether a specific gap clears that margin.
An IQ of 120 on an SD 15 scale sits at the 90.9th percentile, so about 1 in 11 people score that high or higher. It falls in the above-average band. It is a strong result, but the confidence range around it matters as much as the number itself. The IQ comparison tool shows when a gap clears that range.
The average IQ is 100 by definition. Tests are standardised so that the mean of the population is set to 100, and they are normed within age bands, which means the average stays at 100 at every age. About two thirds of people score between 85 and 115. See IQ and age for what actually changes across a lifetime.
The number a test reports does not systematically change, because tests are age-normed — every age band is re-centred on 100. What does change is raw performance: fluid reasoning drifts down from about age 30, while crystallised knowledge keeps rising into your sixties. Any “average IQ by age” table showing different averages per age is measuring something other than IQ. The IQ and age tool plots both.
The score is valid for the year it was taken, but the norms behind it go stale. Population performance drifts upward, so an older test compares you with a lower-performing reference group and returns a higher number. A test normed in 1990 typically reads about 8 points high on today’s scale. Old Test, New Norms converts it, and shows the honest range because the drift rate is disputed. IQ over time walks through the adjustment step by step.
Usually closer to average than you are — in both directions. Two parents averaging 130 have children centring on about 118, and two averaging 70 have children centring on about 82. This is regression to the mean, and the honest answer is a range roughly fifty points wide rather than a single number. See why.
Yes. All seven calculators run entirely in your browser, are free to use, and need no account. Nothing you enter is sent anywhere. Our Security and Privacy Policy covers how the rest of the site handles data.
It depends on the question you actually have. To find out what a score means, start with the IQ percentile calculator — the percentile is the one figure that reads the same on every scale. If your score came from a test using SD 16 or Cattell III B, run it through the IQ score converter first so you are reading the right ruler.
Yes. Every tool here works on any score from any properly normed test — nothing is gated and no account is needed. They are calculators, not a sales funnel. What they cannot do is tell you whether the number you typed came from a real assessment; if you are not sure it did, the IQ test gives you one that states its scale, percentile and confidence range up front.
No. Every calculation runs in your browser — nothing you enter is transmitted, logged or saved, and no account is required. Our Security and Privacy Policy sets out how the rest of the site handles data.
Everything above works on a number you already have. To measure one properly, take the assessment — your score, percentile and full performance report are minutes away.