Realistic random names.
By gender, by country.
First, last, full names — drawn from real public registries, recombined for fiction, demos, and mock data. No signup, instant results, free forever.
About the data
Names are drawn from public-domain sources: US Social Security Administration first-name registry, US Census Bureau surname tables, and equivalent registries for the UK, Spain, Japan, and India. Each pool is trimmed to common entries and shuffled at runtime.
Combinations are random, so any "Maria Tanaka" you see is coincidence. Don't use generated names where claiming a real identity matters.
About this generator
Generated names should sound real without being anyone in particular. That sounds easy until you try it: the synthetic-name generators that ship with most QA tools produce things like "Squiddly McTrout" because they Markov-chain a single English baby-name list. The output is funny once and useless every other time.
This generator does the boring, correct thing instead: real first names from public registries, real surnames from public registries, recombined uniformly at random. The first names come from places like the US Social Security Administration's annual top-1000 list, the UK Office for National Statistics, Spain's Instituto Nacional de Estadística, Japan's Meiji Yasuda registry, and India's voter-roll-derived baby-name compilations. Surnames come from the US Census Bureau's decennial top-surnames release and equivalent registry data abroad. Each pool is the most common 30 to 60 entries — enough variety that 50 draws won't repeat, narrow enough that nothing weird shows up.
The combinations are deliberately uniform. Some name generators try to be "smart" by matching a Mexican first name to a Mexican surname only when you pick "Mexico" — which is reasonable for fiction but wrong for, say, modeling a US customer base, where María González and Sarah Tanaka are equally plausible. Pick "Global mix" when you want diversity; pick a specific origin when you don't. Both modes are honest about what they are.
What this is not: a directory of real people. Every output is a coin flip across two pools, so you'll occasionally get a combination that happens to be someone's real name. That's fine for fiction, NPCs, mock data, and class examples. It's not fine for impersonation, fraud, or any context where a name needs to be verifiably yours. If you need a single random pick from your own list (a class roster, a Slack channel, a contestant pool), the Choice Picker is what you want. If you need to assign people to teams, head over to the Team Picker. If you need a handle rather than a real name, the Username Generator is the right tool.
What people use it for
The boring useful cases — and a couple of fun ones.
Naming fictional characters
Writers, screenwriters, and game designers running a quick batch of side characters. Generate fifty, scroll until something clicks, paste it into the manuscript.
Game testing accounts
Need ten test users that don't all start with "Test"? Generate full names, paste into your QA tooling, and your dashboards stop being a sea of TestUser01.
Mock data for QA
Form fixtures, seed databases, screenshots for design reviews. Realistic names beat Lorem Ipsum because they expose actual layout issues — long surnames, accented characters, names that don't fit one column.
Role-play & tabletop
D&D NPC walks into a tavern. The DM doesn't have a name ready. Twenty seconds later: "Ah, Hiroshi the cooper. Pleased to meet you."
Classroom examples
Teaching SQL, JavaScript, statistics? Generated names beat the same six "Alice / Bob / Carol" placeholders you've used since 2014. They feel like a real dataset because they look like one.
Naming the second cat
People do this. We don't judge. The first-only mode is good for it.
How it works
The data
For each origin we keep two arrays — first names and last names — embedded inline in the page. Each first-name array is split by registry-reported gender and into a curated "neutral" subset for names that appear strongly in multiple categories. There's no network call to fetch any of this; the page ships with the data.
The picking
For each output position we draw a random index using crypto.getRandomValues with rejection sampling — floor(2³² / N) × N as the cutoff so every entry is exactly 1/N probable, with no modulo bias. We dedupe full-name combinations within a single batch (Fisher-Yates over the candidate set) and fall back to allowing repeats only if the request size exceeds the unique-pair count.
State & sharing
Origin, gender, output style, and count live in the URL query string and rehydrate on load. Reload to redraw. The generated list itself is not in the URL — that would make every result a 5KB query string. Recent batches stay in localStorage on this device only, capped at ten.
Common questions
Where do these names come from?
They're drawn from public-domain census and registry data — US Social Security Administration first names, US Census Bureau surnames, and equivalent registries in the UK, Spain, Japan, and India. We've trimmed each pool to the most common 30 to 60 entries per category.
Are these real people?
The first and last names are real, but they're combined randomly. So "Maria Tanaka" might exist, or might not — the generator doesn't know. For fiction, demos, and mock data this is fine. Don't use generated names for any context where claiming a real identity matters.
Can I get just first names or just last names?
Yes. Switch the output style to first only, last only, or first + last (the default). The list updates instantly. Useful when you only need a list of last names for a class roster, or first names for an icebreaker.
How is gender handled?
Pick any to draw from the full pool, or filter to female, male, or neutral. Neutral is a curated subset — names common to multiple genders in the source registries. Surnames are gender-agnostic in every locale we support, so there's no surname filter.
Will the same name repeat in one batch?
Within a single generated batch, full names are deduplicated. If you ask for 50 names from a pool of fewer than 50 unique combinations, the batch will fall back to allowing repeats. The pools are large enough that you'll see this only with extreme combos.
Can I share a configuration?
Yes. The URL reflects every setting — country, gender, style, count. Copy the URL bar and the page rebuilds with the same configuration anywhere. The generated names themselves are not in the URL; reload to draw fresh.
Can I add my own name list?
Not in this tool, but the Choice Picker and List Randomizer handle arbitrary custom lists — paste names, get a random pick or shuffle. For new locales you'd like supported here, request them on the feedback page.
Related generators
Same neighborhood — different question.