*** September 2026 Major Release ***Read and watch
Document type

Extracting data from property tax bills

Each county treasurer or tax collector issues property tax bills in its own design, once or twice a year. Mortgage servicers, underwriters and title companies read one for every escrowed or closing property. It arrives as a PDF from the county site or a scan of the mailed bill.

Servicing needs the parcel number, the assessed value, the full year amount and each installment with its due date. The escrow system schedules those payments, and the annual escrow analysis uses them to set the borrower's monthly deposit.

Page 1 of a sample Property Tax Bill, with extracted values outlined
3,260Markdown characters
106parse blocks
62fields extracted
The problem

What makes property tax bills hard

Every county has its own bill

Each county treasurer designs its own statement, with lettered panels, value tables and a payment coupon arranged in its own order.

Levy rates to many decimals

Levy rates can print to ten decimal places, a precision the tax arithmetic depends on and that a rounded number would lose.

Two addresses on one bill

The property address sits in the statement and the owner's mailing address on the coupon, and for a rental the two are different places.

How it works

Parse, extract, verify

1

Parse

Parse turns the document into reading-order Markdown and a hierarchy of blocks (text, tables, figures), and grounds each block to its exact location on the page.

2

Extract

You define the fields in a JSON schema, and Extract returns their values in a consistent, structured format, each linked back to where it appears in the Markdown.

3

Verify

Each value links back to where it appears, down to the character range in the Markdown, and the page and bounding box in the source document, so every result can be verified.

Worked example

One property tax bill, end to end

About this document: The Clark County, Washington, Treasurer's sample property tax statement, linked from the county's guide to reading a tax bill. One portrait page in lettered sections: header, property and value information, account summary, past-due details, the tax detail by district, and a payment coupon. Page 1 is featured. The owner and addresses are specimen data.

Processing objective: The objective is to follow how the bill is computed. The assessed value times the levy rate gives the district taxes, school district first. Flat assessments bring the total to the full year amount, which is payable in two halves.

Demonstration fields: Assessed value, Levy rate, School district tax, Full year due, First half due

The document

1 page
  • Page 1 of the sample Property Tax BillPage 1shown below

Where each value came from

Page 1 of the sample Property Tax Bill, with every featured field outlined
Assessed valuevalue_information[1].assessed_value
756,979
The region of the sample Property Tax Bill containing Assessed value, with the extracted value outlined
extraction_metadata["value_information[1].assessed_value"]
{
  "value": 756979,
  "page": 1,
  "range": {
    "end": 1297,
    "start": 1289
  },
  "box": {
    "xmin": 0.14337,
    "ymin": 0.56296,
    "xmax": 0.24898,
    "ymax": 0.57843
  }
}
Levy ratevalue_information[1].levy_rate
6.4423487853
The region of the sample Property Tax Bill containing Levy rate, with the extracted value outlined
extraction_metadata["value_information[1].levy_rate"]
{
  "value": "6.4423487853",
  "page": 1,
  "range": {
    "end": 1332,
    "start": 1320
  },
  "box": {
    "xmin": 0.34078,
    "ymin": 0.56296,
    "xmax": 0.47441,
    "ymax": 0.57833
  }
}
School district taxdistrict_taxes[0].tax_amount
345.47
The region of the sample Property Tax Bill containing School district tax, with the extracted value outlined
extraction_metadata["district_taxes[0].tax_amount"]
{
  "value": 345.47,
  "page": 1,
  "range": {
    "end": 2031,
    "start": 2024
  },
  "box": {
    "xmin": 0.7704,
    "ymin": 0.42064,
    "xmax": 0.95215,
    "ymax": 0.43503
  }
}
Full year dueamounts_due.full_year_due
5,157.93
The region of the sample Property Tax Bill containing Full year due, with the extracted value outlined
extraction_metadata["amounts_due.full_year_due"]
{
  "value": 5157.93,
  "page": 1,
  "range": {
    "end": 479,
    "start": 470
  },
  "box": {
    "xmin": 0.7082,
    "ymin": 0.23541,
    "xmax": 0.95119,
    "ymax": 0.25036
  }
}
First half dueamounts_due.first_half_due
2,579.01
The region of the sample Property Tax Bill containing First half due, with the extracted value outlined
extraction_metadata["amounts_due.first_half_due"]
{
  "value": 2579.01,
  "page": 1,
  "range": {
    "end": 418,
    "start": 409
  },
  "box": {
    "xmin": 0.70805,
    "ymin": 0.22063,
    "xmax": 0.95112,
    "ymax": 0.23544
  }
}

How to try it yourself

Everything on this page is reproducible. The sample, its schema, and the Parse and Extract API outputs are published. You can repeat the same two calls from whichever surface you already build on.

  • Claude Skilldescribe the job; the agent writes it
  • CLIno code at all
  • MCPexpose ADE docs as an agent tool
  • REST APIany language
  • Pythonlandingai-ade
  • TypeScriptlandingai-ade
Why LandingAI?

Documents into data

Accurate

The best accuracy on the difficult documents other tools get wrong, such as messy scans, complicated tables, checkboxes and handwriting.

Auditable

Every answer cites its sources. Values come back with an audit trail (page, bounding box, character range) so a human can review them.

Agentic

Applies multi-step reasoning, and plans before it acts. Handles documents it has never seen before and adapts to variations.

API-first

Composable APIs that drop into any tech stack and can be called from anywhere. Easy to build on and embed in your pipelines and products.

Affordable

Cost-competitive rates with complete price transparency. Spend controls before you run, and every job reports its exact cost.

Enterprise-grade

Built for production and scale from day one. Fully observable and monitored, running securely in our cloud or yours.

Built for high-volume, high-throughput production

  • Zero data retention
  • Priority rate limits
  • Enhanced support
  • VPC and on-prem deployment

Available on Team and Enterprise plans.

Compare plans
Capabilities

Property Tax Bill processing highlights

Assessed value
Levy rates at full precision
District tax rows
Special assessments
Full year and installments
Past-due amounts
Parcel and account numbers
Property and mailing addresses
Payment coupon values
Dates normalized to ISO
No per-county templates
Bounding box for every value
FAQs

Questions, answered

Yes. Type the levy rate as a string in your schema, and every printed decimal place comes back intact. Amounts typed as numbers still come back as numbers for the arithmetic around them.

Yes. Define the districts as an array in your schema. Each district row comes back as its own entry with its tax amount, grounded to its own cell in the tax detail table.

No. A schema names the fields you want and describes them in plain language. Extraction reads the parsed bill for meaning, so one schema covers each county's own layout and section order.

Every extracted value carries the character range it was quoted from in the parsed Markdown. The Ground API maps that range to a page and a bounding box, so a reviewer sees each amount in the cell that printed it.

Start extracting property tax bill data

Reliable, structured outputs with full traceability in minutes