Short, medium, and long names

The name generator has a length filter with three settings: short, medium, and long. This guide states how those classes are defined, then reports what the committed SSA data shows when every name in the 2025 top 1,000 for each sex, every decade snapshot from 1955 to 2015, and all 52 per-letter lists are sorted into them. Every figure is computed at build time from the same files the generator samples, through the same functions it calls.

How the three classes are defined

The methodology page defines length on the SSA spelling of the name: short is 4 letters or fewer, medium is 5–7 letters, and long is 8 letters or more. The boundaries printed here are read back from the lengthClassOf function the generator calls — this page probes it with strings of every length rather than retyping the rule — so the classes in this guide cannot differ from the filter.

"Letters" means characters in the SSA's first-name field, and two of the SSA's own rules shape what that field holds. Hyphens and spaces are removed before counting, so Julie-Anne, Julie Anne, and Julieanne arrive as a single entry of 9 letters, which classes as long. Records are restricted to given names of at least two characters, so nothing in the data is shorter than two letters. Spelling variants are separate entries, each classed on its own letters. The provenance document's example of that rule is Caitlin (7 letters, medium), Kaitlyn (7 letters, medium) and Katelyn (7 letters, medium), each spelling with its own rank. Where each spelling sits in the committed files: Caitlin for girls is outside the 2025 top 1,000 and the 2025 C letter list, but charts in the 1985, 1995 and 2005 snapshot top 200 (best #77 in 1985); Kaitlyn for girls is #802 in the 2025 top 1,000 and on the 2025 K letter list, and charts in the 1995, 2005 and 2015 snapshot top 200 (best #34 in 2005); Katelyn for girls is outside the 2025 top 1,000 and the 2025 K letter list, but charts in the 1985, 1995 and 2005 snapshot top 200 (best #59 in 1995). None of the 3 appears in any committed boys list. Absence from a file means absence from that file's committed depth, not from the SSA data.

The 2025 top 1,000, sorted by length

Of the 1,000 girl names in the 2025 top 1,000, 146 are short, 723 are medium, and 131 are long — 14.6%, 72.3%, and 13.1% of the names. Weighted by births the split is 18.5%, 68.4%, and 13.1%: the 146 short girl names account for 219,240 of the 1,182,238 registered births in the list, the medium names for 808,339, and the long names for 154,659.

For boys, the 2025 top 1,000 splits 183 short, 739 medium, and 78 long — 18.3%, 73.9%, and 7.8% of the names, and 20.9%, 69.9%, and 9.3% of the list's 1,387,686 births. The most common exact length in 2025 is 6 letters for girls (291 names) and 6 letters for boys (301 names); girl names in the list run from 3 to 10 letters and boy names from 2 to 11.

The same sort applied nearer the top of the 2025 lists: inside the top 100, girls have 23 short, 67 medium, and 10 long names, and boys 20, 68, and 12; inside the top 10, girls have 2 short, 6 medium, and 2 long, and boys 2, 7, and 1.

The longest and shortest names on the 2025 lists

The longest girl names in the 2025 top 1,000 have 10 letters, and 6 names reach that length; in rank order they are Evangeline (#147, 1,998 babies), Alessandra (#443, 697 babies), Clementine (#456, 676 babies), Alexandria (#474, 647 babies), Jacqueline (#558, 541 babies), Persephone (#735, 372 babies). The longest boy names have 11 letters, reached by 2 names: Christopher (#68, 4,748 babies), Maximiliano (#280, 1,209 babies).

At the other end, the shortest girl names in the 2025 list have 3 letters — 26 names, led by Mia (#6, 11,078 babies), Ava (#11, 7,732 babies), Zoe (#31, 5,646 babies), Ivy (#39, 5,117 babies), Eva (#134, 2,154 babies), Lia (#200, 1,526 babies), and 20 more down to Tru (#974, 262 babies). The shortest boy names have 2 letters: 2 names, Bo (#449, 698 babies), Ty (#813, 306 babies).

These are extremes of the committed depth only. The latest file carries the top 1,000 per sex, so a name registered in 2025 below rank 1,000 — of any length — is outside that file's depth. The only 2025 names past rank 1,000 this site carries are those on a per-letter top-50 list, and those lists were checked too: 526 girl names on the 2025 letter lists rank past 1,000, and the longest of them have 10 letters — Bernadette (#1,034, 241 babies), Wilhelmina (#1,517, 141 babies), Temperance (#2,034, 94 babies), Quinnleigh (#14,521, 6 babies) — the same maximum as the top 1,000; 494 boy names on the 2025 letter lists rank past 1,000, and the longest of them has 11 letters — Quantavious (#11,675, 6 babies) — the same maximum as the top 1,000. Taking the top 1,000 and all 52 letter lists together, the longest girl name anywhere on this site has 10 letters and the longest boy name 11; past those lists nothing is committed, so a longer name may exist in the SSA file without appearing here. The SSA's own rules bound the lists further: names given fewer than 5 times in a year are excluded, and the two-character minimum rules out one-letter entries.

Length across the decade snapshots

The generator's decade eras are single-year snapshots — 1955, 1965, 1975, 1985, 1995, 2005, 2015 — each carrying the top 200 names per sex. A 200-deep list is not comparable with a 1,000-deep one, so the tables below put the 2025 top 200 alongside the snapshots at equal depth. Each row also names the longest and shortest entry in that list (the best-ranked one where several tie) with its letter count.

Girls

Boys

Across the 7 snapshots, the girls' top 200 holds the most short names in 1955 (35) and the fewest in 1995 (16); it holds the most long names in 1995 (40) and the fewest in 1965 (26). The 2025 top 200 for girls has 37 short and 26 long. For boys, the short count is highest in 1965 (51) and lowest in 2005 (34); the long count is highest in 1975 (20) and lowest in 1965 (13); the 2025 top 200 has 43 short and 18 long. These are counts in 7 single years at a fixed depth; they say nothing about the years between snapshots or about ranks past 200 in any of them.

Length by starting letter

Each of the 52 letter pages lists the top 50 names for one initial in 2025 (fewer where fewer qualify: Q and U for girls, Q for boys). Sorting those lists by class: for girls, the letter whose list holds the most long names is G (18 of 50) and the most short names I (21 of 50); for boys, the most long names are under F (11 of 50) and the most short names under N (20 of 50). Letters whose 2025 girl list has no long name at all: U and X; no short name: none. For boys, no long name: none; no short name: none.

How the generator's length filter uses the classes

Choosing a length in the generator keeps only the names whose class matches. The letter, length, and popularity-tier filters apply together, and the sampler draws from whatever survives. With any letter and any tier in the 2025 era, the pools are exactly the counts above: 146 short, 723 medium, and 131 long girl names, and 183, 739, and 78 for boys. Adding the top-10 tier shrinks them: girls 2 short (Emma, Mia), 6 medium (Olivia, Amelia, Sophia, Evelyn, Sofia, Eliana), 2 long (Charlotte, Isabella); boys 2 short (Liam, Noah), 7 medium (Oliver, Henry, James, Elijah, Mateo, William, Lucas), 1 long (Theodore). Adding a starting letter cuts further: girl names starting with G, long, any tier, give a pool of 11 (18 on the G letter list, 7 of them ranked past 1,000); boy names starting with N, short, any tier, give 8 (20 on the N letter list, 12 of them ranked past 1,000). Where a pool and the letter-list count differ, the cause is depth: the generator samples the 2025 top 1,000, while a letter page lists the top 50 names for its initial whatever their overall rank, so a letter-list name ranked past 1,000 is on the page but not in the pool, and an initial with more than 50 names in the top 1,000 has pool names its page does not list. In a decade era the pool for a length is the snapshot count in the tables above, because the era's list is that snapshot's top 200.

Some combinations match nothing, and the generator reports that rather than substituting names. For example, boys in the 2000s (2005 snapshot) era with the short length and the top-10 tier match no name at all in the 2005 data. Pools are never padded: when fewer names match than you asked for, you get all of them, shuffled.

The draw itself is uniform. For the 2025 short girl pool of 146 names, the sampler needs 8 bits to cover the pool, reads 1 byte of randomness, and rejects any value from 146 to 255 — 110 of the 256 possible values — before drawing again, so each of the 146 names is exactly equally likely. The sampling code accepts an injected byte source for testing: fed only zeros, sampleIndices returns positions 0, 1, 2 and the draw is Emma, Mia, Ava — the pool in rank order. In the browser the bytes come from crypto.getRandomValues. The generator sampling guide walks through the full method.

Frequently asked questions

More in the guides index; the length classes, tiers, and depth of every committed list are defined on the methodology page.

Source: every name, rank, and count on this page comes from the U.S. Social Security Administration's national baby-names dataset (1955, 1965, 1975, 1985, 1995, 2005, 2015 and 2025 data, from Social Security card applications). It is public-domain U.S. government data; names given fewer than 5 times in a year are excluded by the SSA, spelling variants are counted separately, and the data covers U.S. births only. Full caveats on the methodology page.