Address parser API that splits any US address into fields
One REST call breaks a raw address line into house number, street, unit, city, state, and ZIP. Built for developers.
Got a whole list? Upload the file and clean every row →
What comes back
Each parse returns the address split into clean components, plus the match metadata to trust it.
What it reads and cleans up
Raw address strings are messy. Here is what the parser resolves, with real examples.
Misspelled streets and cities
Missing state or ZIP
Abbreviations and unit types
Out-of-order lines
Measured, not claimed: on 26,943 corrupted real US addresses (August 2026 benchmark) the parser recovered the right address 96.2% of the time on single typos or dropped fields and 99.9% on formatting-only damage, and refused to guess 99.9% of the time when the house number did not exist. Full results by damage type in the accuracy benchmark; the test method is published so you can run it on any vendor.
Common use cases
A single GET request
Send a raw address, get structured JSON back. No SDK required.
# raw input, one call curl "https://api.sthan.io/v2/address-parser/usa/speculative/1600%20pennsylvania%20ave%20nw%20washington%20dc%2020500" \ -H "Authorization: Bearer YOUR_API_KEY"
{
"inputAddress": "1600 pennsylvania ave nw washington dc 20500",
"fullAddress": "1600 Pennsylvania Ave NW, Washington, DC 20500-0005",
"addressLine1": "1600 Pennsylvania Ave NW",
"addressLine2": "Washington, DC 20500-0005",
"addressNumber": "1600",
"streetPreDir": null,
"streetPreType": null,
"streetName": "Pennsylvania",
"streetPostType": "Ave",
"streetPostDir": "NW",
"unitType": null,
"unitNumber": null,
"city": "Washington",
"stateCode": "DC",
"state": "District Of Columbia",
"zipCode": "20500",
"zip4": "0005",
"county": "District Of Columbia",
"matchMode": "Speculative",
"matchTier": "Exact",
"confidence": 1,
"matchCode": {
"houseNumber": "Matched",
"street": "Matched",
"unit": "NotApplicable",
"city": "Matched",
"state": "Matched",
"zipCode": "Matched",
"zip4": "Inferred"
},
"matchedFromCache": false,
"isError": false
}
Start free, scale when you need to
Monthly plans, billed by request volume. Annual saves up to 50%.
Higher volumes and annual discounts (up to 50%) on the plans page.
Questions developers ask
What is address parsing?
Breaking a single address string into its parts: house number, street name and type, unit, city, state, ZIP and ZIP+4, so you can store, sort, and integrate structured fields.
How is this different from verification?
The parser splits a raw string into fields. Verification also checks the address against postal data and returns a deliverable status. Many teams parse first, then verify.
Does it handle messy input?
Yes. It reads typos, abbreviations, missing parts, and out-of-order lines, and returns clean structured fields instead of an empty result.
How fast is a parse?
Complete addresses typically return in 1 to 5 seconds. Partial or ambiguous input can take longer while it applies deeper recovery.
What do I need to get started?
An email address. Sign up, grab your API key, and send your first request. The free tier gives you 3,000 requests a month (up to 100 a day) with no card on file.
Parse your first address in two minutes
Sign up, copy your key, make the call. 3,000 requests a month, free.
Get your free API key