Frankfurter API

Frankfurter is a free, open-source exchange rate API. The public API lives at https://api.frankfurter.dev — no API key, no signup, no monthly quota (only abuse-prevention rate limiting). It tracks daily reference rates from 84 central banks covering 201 currencies, with history back to 1948.

Use the v2 API (/v2/...) documented below. The older v1 API still works; see references/v1-api.md only if you must maintain existing v1 code.

Quick reference

Base URL: https://api.frankfurter.dev

Currency codes are ISO 4217 (USD, EUR, TWD, JPY...). Dates are YYYY-MM-DD.

Response shape

/v2/rates returns a JSON array of rows; /v2/rate/{BASE}/{QUOTE} returns a single object:

// GET /v2/rates?base=USD&quotes=EUR,TWD
[
  {"date": "2026-07-03", "base": "USD", "quote": "EUR", "rate": 0.87598},
  {"date": "2026-07-03", "base": "USD", "quote": "TWD", "rate": 31.913}
]

// GET /v2/rate/EUR/USD
{"date": "2026-07-03", "base": "EUR", "quote": "USD", "rate": 1.1416}

A time series is the same array with one row per date per quote currency.

Currency conversion

There is no conversion endpoint in v2. Fetch the pair rate and multiply:

import requests

def convert(base: str, quote: str, amount: float, date: str | None = None) -> float:
    url = f"https://api.frankfurter.dev/v2/rate/{base}/{quote}"
    params = {"date": date} if date else {}
    r = requests.get(url, params=params, timeout=10)
    r.raise_for_status()
    return amount * r.json()["rate"]

print(convert("USD", "TWD", 100))  # e.g. 3191.3

async function convert(base, quote, amount) {
  const r = await fetch(`https://api.frankfurter.dev/v2/rate/${base}/${quote}`);
  const d = await r.json();
  return amount * d.rate;
}

Providers and blended rates

By default, rates are blended across all contributing central banks. This is fine for general use, but the last decimal places can shift as new data arrives.

• For official/compliance use, pin a specific source: ?providers=ECB (provider keys come from /v2/providers — e.g. ECB, BOE, FRED, CBC).
• To see which sources fed a blended rate, add ?expand=providers: each row gains a providers array of {key, rate} objects. Pegged currencies omit it (their rate comes from the peg, not provider data).

Output formats

• JSON (default).
• CSV: append .csv to the path (/v2/rates.csv?...) or send Accept: text/csv. Columns: date,base,quote,rate.
• NDJSON: send Accept: application/x-ndjson. Each line is one independent JSON object — use this to stream large time series without buffering.

Errors

Errors return an HTTP status with a JSON body like {"status": 404, "message": "not found"}:

• 400 — invalid parameter or malformed request
• 404 — currency, rate, or resource not found (check the ISO code and date)
• 422 — request understood but cannot be processed

Practical tips

• The public API blocks the default Python-urllib User-Agent (Cloudflare returns 403). curl, fetch, and the requests library work as-is; if you use urllib, set any custom User-Agent header.
• Rates are daily reference rates, not live tick data. Weekends and holidays have no new data, so "latest" may be the previous business day — read the date field in the response rather than assuming today.
• Narrow quotes to the currencies you need; it keeps responses small and fast.
• Cache responses when polling — data updates at most daily per provider, so requesting more often than hourly is wasted traffic.
• Historical coverage varies by currency: check start_date in /v2/currency/{CODE} before requesting old dates.
• Free for commercial use; see each provider's terms for the underlying data.
• Self-hosting is available via Docker for full control (see frankfurter.dev).

For complete endpoint details with verified request/response examples, read references/endpoints.md.