Back to tools
3 endpoints

Edmunds

New and used car inventory search by make, model, year and location, with specifications and pricing.

Part of Data

Get started by giving this to your agent
$Set up https://monid.ai/SKILL.md, and then use Monid to show me what I can do with Edmunds.
Edmunds
Get vehicle detail
/get_vehicle_details

Search new and used car inventory on Edmunds by make, model, year, condition, and location. Retrieve detailed specs, pricing, history, and dealer info by VIN, and browse all available car makes.

$0.0100per call
0 Markup
Edmunds
Search inventory
/search_inventory

Search new and used car inventory on Edmunds by make, model, year, condition, and location. Retrieve detailed specs, pricing, history, and dealer info by VIN, and browse all available car makes.

$0.0100per call
0 Markup
Edmunds
List makes
/get_makes

Search new and used car inventory on Edmunds by make, model, year, condition, and location. Retrieve detailed specs, pricing, history, and dealer info by VIN, and browse all available car makes.

$0.0100per call
0 Markup

Try it with your agents

01
Ground a decision in census data
Set up https://monid.ai/SKILL.md, and then use Monid to pull census demographics for Austin, Denver, and Raleigh and compare their age and income distributions.

How-to guide

Is this used car priced fairly?comps by hand, and what one word on the report is worthsearch carfax on monid for a <year> <make> <model> within 50 miles of <zip> and give me price, mileage, trim, dealer and the accident summary for each. check the year on every row yourself: this provider's year filter leaks badly, about a third of the rows survived it on both markets we tried. take the service-record count from the number inside the service-history object, not from the similarly named field next to it, which is a boolean what one row holds: price, mileage, trim, dealer, distance, owner count, accident summary, service-record count and the fees, so 25 cars cost $0.0004 each to shortlist the filter leaks: a search pinned to one model year reported 115 results of which 42 were that year, and pages three and four held none of it; on a second market 12 of 38 rows were the year asked for and the last page held none either but it is this provider, not the aisle: the second source's year filter was clean on a second market, 17 rows of 17 on-year, so do not tell a reader every retail search leaks and the dishonest count is the other one: this provider's total was accurate on the second market, 38 reported and 38 reachable, while the second source returned 17 rows while reporting a total of 8 one field name is a trap: the service-record COUNT lives inside the service-history object; the field named for service records next to it is a boolean, so a reader who reaches for the obvious name gets true rather than a number the ZIP that was ignored: another provider accepted the ZIP, returned 21 cars from seven other states, computed the distances from somewhere else, and billed a 200 for it coverage is the real difference: 115 rows against 30 against 21 for the same search, and the subject car existed in only one of the three