Free Roman numeral API

Endpoint

Each number has its own file. Request it directly and you get back JSON with the numeral and the words.

GET https://romanize.net/api/roman/{number}.json

Example request

curl https://romanize.net/api/roman/2026.json

Example response

{
  "number": 2026,
  "roman": "MMXXVI",
  "words": "two thousand twenty-six",
  "valid": true
}

From JavaScript

const res = await fetch("https://romanize.net/api/roman/49.json");
const data = await res.json();
console.log(data.roman); // "XLIX"

From Python

import urllib.request, json

url = "https://romanize.net/api/roman/49.json"
with urllib.request.urlopen(url) as res:
    data = json.load(res)

print(data["roman"])  # XLIX

From PHP

<?php
$url = "https://romanize.net/api/roman/49.json";
$data = json_decode(file_get_contents($url), true);

echo $data["roman"]; // XLIX

From Ruby

require "net/http"
require "json"

url = URI("https://romanize.net/api/roman/49.json")
data = JSON.parse(Net::HTTP.get(url))

puts data["roman"] # XLIX

Every one of these is a plain GET against a static file. There is no authentication step to add, no SDK to install, and no response shape to negotiate: the file at /api/roman/{n}.json is the whole contract. It is served with Content-Type: application/json and Access-Control-Allow-Origin: *, so a browser fetch from any origin works as well as a server-side one.

Fields

Usage and fair play

The API is served as static files from a content delivery network, so it is fast and resilient, but please cache responses on your side rather than fetching the same number repeatedly. Attribution is appreciated but not required. If you build something with it, we would love to hear about it on the contact page.

Frequently asked questions

The rejected forms, as a second dataset

The API returns what is valid. The other half is published separately: every malformed numeral form with the reason it fails and its canonical correction, as CSV and JSON under CC BY 4.0. See the rejected forms dataset.

Prefer to embed the UI?

If you want the converter itself rather than raw data, use the embeddable widget. Romanize keeps both free.