Random Name Generator — by origin, gender, and style

The generator pairs a randomly drawn first name with a randomly drawn surname, both pulled from origin-specific lists. There are seven origin pools: English, Sinhala (Sri Lanka), Tamil, Spanish, Japanese, a high-fantasy set, and a near-future sci-fi set. The English, Sinhala, Tamil, Spanish, and Japanese lists are derived from public-frequency tables published by census bureaus and registry offices in those countries — sources are cited at the bottom of this page. The fantasy and sci-fi lists were written for this tool and are original.

Each pick goes through three steps. (1) An origin is picked uniformly across the boxes you ticked — equal weight, so checking “English” and “Sinhala” gives you a 50/50 mix regardless of which origin has the longer list. (2) A gender is resolved. In “any” mode the male and female pools of the chosen origin are concatenated, so neither side is over-represented when pool sizes differ (the English pool has 33 male and 33 female names — the weighting still matters for origins like Japanese where pools come out slightly differently). In male or female mode, only that pool is used. (3) A first name is drawn from the resolved pool and a surname from the (gender-neutral) surname list of the same origin. For username and email formats, an extra random draw appends a 2-digit numeric suffix or strips the name to ASCII for [email protected]form.

Randomness flows from a 32-bit Mulberry32 PRNG (public domain, by Tommy Ettinger). The PRNG is seeded with an integer that's drawn fresh from the browser's Web Crypto API (crypto.getRandomValues) when the page loads, so each new batch is cryptographically unpredictable. The seed itself is then shown in the UI and round-trips through the URL — clicking “Copy link” adds ?seed=… to the address, and anyone opening it sees your exact batch. This design is the standard trick for free-tier name pickers: the unpredictable choice is the seed; everything downstream is deterministic so it can be reproduced for screenshots, code reviews, or test fixtures.

The “name space” figure on the tool shows N = Σ (firstPool × surnamePool) across the origins you ticked, and the entropy figure is log₂(N) — the number of bits an attacker would need to guess to predict your next single name. With all seven origins on and gender = any, the space is 21,114 unique full-name combinations across 508 first names and 292 surnames. That's about 14.4 bits per draw — plenty for fiction, test data, or character casts, but a long way short of what you'd use for a password or token.

The tool runs entirely in your browser. There is no server-side endpoint that sees your choices or the names, no analytics on the generated values, and the entire wordlist ships as part of the page bundle. You can disconnect from the internet after the page loads and the generator still works.

• US Social Security Administration — top baby names (public domain)
• UK Office for National Statistics — baby names England & Wales
• US Census Bureau — 2010 frequently-occurring surnames
• Department of Census & Statistics, Sri Lanka — population name frequencies
• Instituto Nacional de Estadística (Spain) — most frequent names and surnames
• Japan Ministry of Justice — baby-name registry data
• MDN — Crypto.getRandomValues() (used to seed the PRNG)
• Mulberry32 PRNG — Tommy Ettinger (public domain)

The seven origin corpora were last reviewed against the cited frequency tables on 2026-05-11. The fantasy and sci-fi lists are original copy edited at the same time. Lists are reviewed annually or when a cited source publishes a refreshed frequency table.