Extracting data from vehicle window stickers
A manufacturer fixes a window sticker, the Monroney label, to every new car before it reaches the dealer, and each brand designs its own layout. Auto lenders and insurers receive one with each new-vehicle deal they fund or cover, as a phone photo of the glass or a PDF reprint pulled by VIN.
An underwriter needs the base MSRP, each priced option, the destination charge, the total MSRP and the VIN. A loan origination system can cap the advance at a share of that total. An insurer's rating system uses it to value a newly bought car.
What makes vehicle window stickers hard
Each manufacturer prints its own layout
Federal law sets what the label must show, and each manufacturer decides where. Option lists and panel order change from brand to brand.
Four regulated panels share one sheet
One label carries the price list, the EPA fuel economy label with its rating sliders, the NHTSA star ratings and the parts content statement.
The label stays on the glass
The law keeps the sticker on the car until the buyer takes delivery, so copies often arrive as phone photos with glare and skew.
Parse, extract, verify
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.
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.
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.
One vehicle window sticker, end to end
About this document: The Monroney label of a Kia Niro EX Touring on an auto show display car, one landscape image cropped from a phone photo. It carries five panels: standard features, the price column, the EPA fuel economy label, the NHTSA safety ratings and the parts content statement. Ceiling lights glare off the display glass. The EPA label prints its combined MPG in large numerals beside a fuel pump icon.
Processing objective: The objective is to capture what the car costs to buy and to run, for a lender's loan-to-value check and an insurer's valuation. The price column reconciles: the 31,490.00 base MSRP, 480.00 in priced options and the 1,325.00 freight charge make the 33,295.00 total.
Demonstration fields: Base MSRP, Inland freight and handling, Total MSRP, Combined MPG, Annual fuel cost
The document
1 page
Page 1shown below
Where each value came from
pricing.base_msrp
{
"value": 31490,
"page": 1,
"range": {
"end": 1932,
"start": 1921
},
"box": {
"xmin": 0.50599,
"ymin": 0.15368,
"xmax": 0.56374,
"ymax": 0.17597
}
}pricing.inland_freight_and_handling
{
"value": 1325,
"page": 1,
"range": {
"end": 3060,
"start": 3051
},
"box": {
"xmin": 0.50597,
"ymin": 0.80883,
"xmax": 0.56394,
"ymax": 0.85666
}
}pricing.total_msrp
{
"value": 33295,
"page": 1,
"range": {
"end": 3141,
"start": 3131
},
"box": {
"xmin": 0.50597,
"ymin": 0.85632,
"xmax": 0.56395,
"ymax": 0.89345
}
}fuel_economy.combined_mpg
{
"value": 49,
"page": 1,
"range": {
"end": 4162,
"start": 4144
},
"box": {
"xmin": 0.5694,
"ymin": 0.231,
"xmax": 0.66568,
"ymax": 0.30424
}
}fuel_economy.annual_fuel_cost
{
"value": 900,
"page": 1,
"range": {
"end": 4413,
"start": 4409
},
"box": {
"xmin": 0.60023,
"ymin": 0.41775,
"xmax": 0.67788,
"ymax": 0.46911
}
}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
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. VPC and on-prem deployment require an Enterprise plan.
