Free Random Number Generator

JavaScript's Math.random() is a pseudo-random number generator (PRNG), not a true random source. It uses a deterministic algorithm — typically xorshift128+ or a similar design — seeded by engine state at startup. Every call produces a number in [0, 1) by advancing internal state through a fixed mathematical transformation. Given the same starting seed you would get the exact same sequence every time.

For the vast majority of everyday uses — games, simulations, random sampling, picking a winner, rolling dice — this is completely fine. The output passes rigorous statistical randomness tests and is indistinguishable from true randomness for those purposes. The determinism only matters if an adversary can observe enough outputs to reconstruct your seed, which is the threat model of cryptography.

Do not use Math.random() for: passwords, session tokens, API keys, OAuth secrets, cryptographic nonces, or anything where predictability is a security risk. Do use it for: games, dice rolls, random sampling, shuffles, simulations, and test data generation. For security-sensitive values, the browser's crypto.getRandomValues() draws from the OS entropy pool and is cryptographically secure.

A common mistake when mapping random bytes to a custom range is using the modulo operator (%). The problem: modulo produces bias unless your source space divides evenly into your target range. Consider mapping a random byte (0–255, so 256 possible values) to a die face (1–6). You might write result = (byte % 6) + 1. The byte values 0–251 map cleanly across six faces 42 times each. But 252, 253, 254, and 255 map to faces 1, 2, 3, and 4 respectively — giving those four faces one extra hit each. Faces 1–4 appear 43 times per 256 draws; faces 5–6 appear only 42 times. That is modulo bias.

The correct fix is rejection sampling: if the drawn byte falls in the biased tail (here, values ≥ 252), discard it and draw again. Repeat until you get a value in the unbiased portion. Each retry is independent, so the expected number of extra draws is tiny — less than one on average for most practical ranges. The result is that every outcome in your target range has exactly equal probability, with no skew toward the low end.

This tool handles the bias problem internally so you never have to think about it. Every number in your Min–Max range has equal probability regardless of how the range size aligns with the underlying random source.

The obvious way to generate n unique random numbers is to generate one, check if you've seen it before, and retry if you have. This works fine when you need a handful of values from a large range — collisions are rare and retries are cheap. But performance degrades sharply as the count approaches the range size, for the same reason the birthday problem surprises people: once you've filled roughly 50% of the available slots, about half of all draws are collisions, so you're doing roughly two draws per accepted value. At 90% fill, nine out of ten draws are wasted. Near 100%, the retry loop runs for an arbitrarily long time and the behavior becomes unpredictable.

The Fisher-Yates shuffle sidesteps this entirely. It works by maintaining a virtual pool of all available values and swapping each selected value out of the pool permanently. Each draw is O(1) with zero chance of collision because you are always picking from the remaining unchosen values. The full algorithm runs in O(n) time — linear in the count you want — regardless of how close that count is to the range size. There is no retry logic and no worst-case blowup.

This tool uses Fisher-Yates for no-duplicates mode, which is why it handles requests like "999 unique numbers from 1–1000" just as quickly as "5 unique numbers from 1–1000."

1. Set your range. Enter a minimum and maximum value. The generator will produce numbers within this range (inclusive). Both positive and negative numbers are supported.

2. Choose single or bulk mode. Set Count to 1 for a single number, or increase it up to 1,000 for bulk generation. Single mode shows a large, prominent result; bulk mode displays all numbers in a compact grid.

3. Configure options. Toggle "No Duplicates" to ensure unique numbers. Choose ascending, descending, or unsorted order. The tool remembers your preferences between visits.

4. Roll dice. Use the dice roller section to simulate standard RPG dice. Click any die button to roll it, and view your complete roll history. Copy or clear results anytime.