hmrc-rates

HMRC exchange rates for Python, backed by the hmrc-rates Rust crate. The full published history ships inside the wheel, so Rates() is free and needs no network, parsing or I/O. All money values are decimal.Decimal and conversions are exact — the library never rounds.

Install

python3 -m pip install hmrc-rates

Python ≥ 3.10. Wheels ship for Linux (glibc and musl, x86_64 and aarch64), macOS and Windows. The sdist builds anywhere with a Rust toolchain.

Quick start

import datetime
from decimal import Decimal

from hmrc_rates import YearMonth, Rates, YearEnd

rates = Rates()

# Monthly customs/VAT rate; months can be YearMonth, datetime.date, or "YYYY-MM"
rate = rates.monthly_rate("USD", YearMonth(2025, 8))
gbp = rate.to_gbp(Decimal("2500"))  # exact — you choose the rounding
print(f"$2500 in Aug 2025 = £{gbp.quantize(Decimal('0.01'))}")

# Self Assessment style: the yearly average to 31 March 2025
eur = rates.average(YearEnd.march(2025)).rate("EUR")
print(f"EUR average, year to 31 Mar 2025: {eur.units_per_gbp}")

# Whole tables iterate as (Currency, Rate) pairs
table = rates.monthly(datetime.date(2025, 8, 15))
print(f"{len(table)} currencies in {table.period}")

"GBP" always resolves to the identity rate, and code lookups are case-insensitive.

Data coverage

One Rates value holds all four series HMRC has published:

Currency codes are as published by HMRC, which is not always ISO 4217. E.g. Ecuador appears as ECS.

Strict lookups and exceptions

A period HMRC never published is an exception, never a silently substituted older rate. Every exception derives from HmrcRatesError (not the builtin LookupError):

• UnknownCurrencyError — the code never appears in that series
• PeriodNotAvailableError — no table for that period (the message includes the available range)
• NotInPeriodError — the period exists but the currency is absent from it
• FetchError — fetching fresh rates failed

Fallback is opt-in and visible:

rate = rates.monthly_rate_or_earlier("USD", month, 2)
rate.period  # reveals which month was actually used

Exactness

Amounts must be decimal.Decimal or int: floats raise TypeError because they are inexact. Results come back as decimal.Decimal, unrounded. Apply whatever rounding your tax context requires.

Fresh rates

Updater fetches whatever HMRC has published since this release and caches the files verbatim in the system cache directory. refreshed() performs blocking HTTP with the GIL released. cached() never touches the network.

from hmrc_rates import FetchError, Updater

updater = Updater()
try:
    rates = updater.refreshed()
except FetchError as e:
    print(f"warning: possibly stale rates: {e}")
    rates = updater.cached()  # offline fallback is an explicit choice

New HMRC publications also trigger a new patch release of this package automatically. So pinning the latest version is an alternative to runtime fetching.

Development

cd python
uv venv && uv pip install maturin pytest
VIRTUAL_ENV=$PWD/.venv .venv/bin/maturin develop
.venv/bin/pytest

The package is a PyO3 binding built with Maturin. The Rust sources live in the repository root.

Licence

MIT.

Disclaimer

This project is not affiliated with, endorsed by, or maintained by HMRC. Only the underlying rate data comes from HMRC.