datamade/usaddress

Health v2 · maintenance only

Flags: no_releases

Adoption not part of the score

1636 stars · 308 forks observed · 2026-08-28

What it is AI-extracted, prompt v1, taxonomy v1, 2026-08-30, confidence not recorded

usaddress is a Python library that parses unstructured United States address strings into labeled address components using a conditional random fields model. It provides parse and tag methods that split messy addresses into fields like AddressNumber, StreetName, and PlaceName, without validating or normalizing addresses.

Use cases

• parse unstructured US address strings into components
• split addresses into street, city, state, and zip columns
• extract address fields from messy data for cleaning
• tag address parts in a dataset before geocoding
• handle tricky addresses that rule-based parsers fail on

When to choose

• you need to parse free-form US addresses in Python
• your addresses are messy and rule-based parsers break down
• you want an MIT-licensed, pip-installable address parser
• you need labeled address components as input to geocoding or deduplication

When to avoid

• you need address validation or deliverability checks
• you need address normalization or standardization out of the box
• you need to parse non-US addresses
• you require perfect accuracy - the model is probabilistic

Facets

library · maturity active

nlp parser machine-learning developer-tools data-science python cross-platform address-parsing conditional-random-fields us-addresses geocoding-preprocessing probabilistic-parsing natural-language-processing

Member repositories

For agents

markdown · JSON · MCP: product_card(name="datamade/usaddress")