Onomaverse Names Datasets

Open data on 38,069 names across 106 countries and 94 languages: name frequencies, multilingual transliterations, gender inference, a name-equivalence graph, and name-day calendars.

🔗 Source & full documentation: onomaverse.com/datasets 📄 License: CC BY 4.0 · 🆔 DOI: 10.5281/zenodo.20797446

Names data from Onomaverse, licensed CC BY 4.0.

Subsets

Each subset is a separate config — load any of them by name. Per-subset documentation lives on the website.

Subset (config_name)	Rows	What it is	Page
given-name-frequency	53,868	First-name frequency per country, with gender & in-country share	page
surname-frequency	41,248	Surname frequency per country, with in-country share	page
name-gender-inference	18,208	Male/female counts and P(male) per first name	page
name-transliterations	1,126,144	Each name across up to 94 languages & scripts	page
name-equivalence	70,626	Variants & cross-language equivalents (John ↔ Juan ↔ Ivan)	page
name-days	2,426	Name-day dates by name & region (CSV + importable .ics)	page
popular-names-by-country-2026	15,654	Top 100 given names & surnames per country	page

Quick start

from datasets import load_dataset

# Pick any subset by its config name:
ds = load_dataset("onomaverse/names", "name-transliterations", split="train")
print(ds[0])

gnf = load_dataset("onomaverse/names", "given-name-frequency", split="train")

Every subset also ships as CSV / JSON Lines / Parquet in this repo, so you can grab files directly:

import pandas as pd
df = pd.read_parquet("hf://datasets/onomaverse/names/given-name-frequency.parquet")

Attribution (required)

Names data from Onomaverse, licensed CC BY 4.0. A link back to onomaverse.com/datasets satisfies the attribution requirement.

This dataset contains no descriptive prose (meanings, etymology, biographies). For that — for every name, in 94 languages — visit onomaverse.com.