Random Number Generator

How to use the generator

1. How many numbers: from 1 to 500.
2. Minimum and maximum: both ends can be drawn. For a die, use 1 and 6.
3. Number type: whole numbers, or decimals with 1 to 10 decimal places.
4. No repeats: each value appears at most once. You can't ask for more numbers than there are possible values.
5. Sort: lists the numbers from smallest to largest.
6. Seed (advanced, optional): any text. Leave it empty for a fresh draw every time.

Below the list, Possible values tells you how many different numbers could come up, and Random source shows which generator was used.

How the numbers are generated

Without a seed

The draw uses your browser's crypto.getRandomValues, which MDN describes as producing cryptographically strong random values. If a browser doesn't offer it, the generator falls back to Math.random, which MDN says does not provide cryptographically secure random numbers. The source line tells you which one ran. Either way, a new list appears on every draw and can't be recreated.

With a seed

The seed text is turned into a number with the FNV-1a hash, which starts a small deterministic generator called mulberry32. The same seed with the same settings always gives the same list, on any device. Spaces around the seed are ignored, but capitals matter: “Class” and “class” give different lists. A seed makes results repeatable and checkable, and also predictable to anyone who knows it.

Fair odds within the range

Each number is built from 53 random bits, and draws that would favor some values are thrown away and redrawn (rejection sampling). That avoids the modulo bias of the common “random % range” shortcut, so every possible value has the same chance. Decimals are drawn from the multiples of 0.01 (for 2 places) between the bounds. In whole-number mode, decimal bounds shrink to the whole numbers inside them: −5.5 to 5.5 gives −5 to 5, or 11 possible values. With no repeats, the generator shuffles the candidates with a Fisher–Yates shuffle (up to 10,000 of them) or redraws duplicates in larger ranges.

Examples

Draws without a seed change every time, so these examples use seeds you can type in to get the same output.

1. Ten numbers from 1 to 100

With seed 42: 50, 67, 18, 16, 78, 99, 70, 77, 17, 69. There are 100 possible values. With seed 43 the list is 64, 33, 37, 6, 81, 67, 38, 57, 57, 43. Notice that 57 comes up twice, which is normal when repeats are allowed.

2. Shuffle a class of 30

Number your students 1 to 30, ask for 30 numbers from 1 to 30 and tick No repeats. With seed class the order begins 25, 18, 29, 11, 5. Every student appears exactly once. Share the seed and anyone can check the order.

3. Decimals between 0 and 1

Five numbers with 2 decimal places, seed 42: 0.44, 0.33, 0.26, 0.07, 0.16. There are 101 possible values, from 0.00 to 1.00.

What not to use it for

Do not use this generator for lotteries, gambling, raffles or contests with legal or financial stakes, passwords, keys, or any security decision. A seeded list is predictable by design, Math.random is not cryptographically secure, and these draws are not certified or audited. For keys and other security uses, NIST publishes standards for random bit generation, and MDN recommends the Web Crypto API's generateKey() for creating keys.