Last name generator

Choose a last name for your character

Start with Generate 12 names for a random selection, or narrow the choices first. Starts with matches the beginning of the recorded spelling; Ends with matches its ending. Use both to set a pattern, such as a name beginning with B and ending in son. Leave either box empty when that part does not matter.

Length counts the letters in the published spelling. The short range is two to five letters, medium is six to eight, and long is nine to fifteen. People in 2010 selects a recorded-count range: 100–999, 1,000–9,999, 10,000 or more, or every published count. All four filters apply together, so a restrictive combination may have few matches or none.

After changing a filter, generate again to apply it. Reset filters returns to the full surname pool and makes another set. It keeps your shortlist. If fewer than twelve spellings match, the tool shows all of them; repeating the request cannot create additional names that the source does not contain.

Keep the names you want to use

Select Shortlist beside any suggestion to keep it while trying more names. You can keep twelve at a time, remove individual names or clear the list. A saved name stays selected if it appears in another set.

Copy names puts one name per line on your clipboard, ready to paste into your notes. If automatic copying is unavailable, the page provides a selected text box. Save list · TXT includes each name's count, Census rank, year and source. Save or copy before refreshing or leaving: the list stays only in this page, with no account or browser storage.

To inspect every matching surname in an ordered list, use the surname explorer. For first-name ideas, the name explorer uses a separate dataset of baby-name records by birth year.

Random spellings and surname frequency

The generator samples spellings equally. A surname reported by 100 people has the same chance of being selected as one reported by a million people, when both match your filters. There are no repeats within one set; a later set can include a previous suggestion. This helps surface unfamiliar options, but the results are not a random sample of the U.S. population.

Every suggestion carries its original recorded count and official Census rank. Filtering does not recalculate that rank. The most common last names page provides a frequency-ordered reference, while rare last names explains the smallest published counts. A low count in 2010 does not mean a name is unique or equally uncommon today.

Where these surnames come from

The source is the U.S. Census Bureau's 2010 surname tabulation: 162,253 spellings reported by at least 100 people each. Counts describe that historical U.S. population. They do not establish ancestry, a name's origin or meaning, or a worldwide total. A surname recorded in the United States is not necessarily English in origin.

The published keys contain only A–Z, with up to fifteen letters. Original accents, punctuation, spaces and capitalization cannot be reconstructed from this file. The generator uses initial capitals for readability; its pattern filters fold accents and punctuation to the source's letters and explain a changed pattern. Unsupported scripts and digits get a correction message instead of a guessed match.

The same public file serves every visitor. Filters, random selection, shortlisting and downloads run in your browser; no typed pattern is sent in a request. The surname methodology documents the publication threshold and source editing rules.