Sign up and add a payment method to enable your API key, send the key as a Bearer token in the Authorization header, then call the endpoint that matches how you identify properties — address, parcel ID, owner name, or location. Responses are JSON. Most first integrations take under an hour.
Realie follows ordinary REST conventions, so integration is mostly about choosing the right lookup path and handling volume sensibly. Pick the endpoint that matches the identifier you already hold: /property/address when you have a street address, /property/parcelId when you have a county APN, /property/owner when you are working from a name, /property/location for a coordinate and radius, and /property/search for filtered queries. Premium endpoints add fuzzy owner matching and comparables. Beyond the first successful call, the work that matters is batching, caching, and respecting the rate limit.
Create an account and enable API access
Sign up, add a payment method in Billing, then open Usage > API access to generate your API key. The free tier is available without a sales call.
Authenticate
Send your key as a Bearer token in the Authorization header: Authorization: Bearer YOUR_API_KEY. Keep the key server-side; never ship it in client-side code.
Choose the endpoint that matches your identifier
Use /property/address for street addresses, /property/parcelId for county APNs, /property/owner for name searches, /property/location for coordinate-and-radius queries, and /property/search for filtered queries.
Make a first request and inspect the schema
Run one lookup against a property you already know and read the full JSON response, so you can map fields deliberately rather than guessing at names.
Batch your requests
A single request can return up to 100 parcels. Batching reduces both latency and token consumption compared with looping one property at a time.
Handle errors and rate limits
Handle non-200 responses explicitly and add retry with backoff so a transient failure or a rate-limit response does not drop records from your pipeline.
Cache what does not change
Parcel geometry and structural characteristics change rarely. Caching them locally and refreshing on the county cycle cuts both cost and response time.
Two decisions determine whether an integration stays cheap as it grows. The first is batching: because one request can carry up to 100 parcels, a loop that fetches properties individually can cost far more than the same workload batched. The second is caching. County records refresh on a county cycle, not continuously, so re-fetching unchanged parcel geometry on every page load spends tokens for no new information.
For terminal and coding-agent workflows there are two alternatives to writing HTTP calls yourself: the Realie CLI, which supports a --json flag for piping into other tools, and an MCP server that connects the data directly to IDE agents like Claude, Cursor, and VS Code.