For developers & AI assistants

Everything on this site is available as a Model Context Protocol server and as a plain REST API. The fitment maths is deterministic — same input, same answer, every time.

MCP endpoint

https://whatfitsmycar.com/mcp

Streamable HTTP transport, no authentication. Add it as a custom connector in Claude or ChatGPT.

Tools

find_tires_for_vehicle

Factory tire and wheel sizes for a year/make/model, including staggered setups.

decode_vin

17-character VIN → vehicle identity via the official NHTSA database, then its tire sizes.

check_tire_fitment

Candidate size vs original: diameter difference, speedometer error, load and speed rating.

find_alternative_sizes

Every manufactured size within ±3% overall diameter, optionally for a different wheel diameter.

explain_tire_size

Decodes a sidewall marking into width, sidewall height, diameter, load kg and speed km/h.

shop_tires

Available tires in a size with prices where merchant data is loaded, plus retailer links.

REST

GET /api/size/225-45R17
GET /api/fitment?original=225/45R17&candidate=235/45R17
GET /api/alternatives?size=225/45R17&rim=18
GET /api/vehicle?make=Honda&model=Civic&year=2019
GET /api/vin/1HGFC2F59KH500000
GET /api/shop?size=225/45R17&region=US

Rate limits & data

60 requests/minute per IP on the website and REST API, 600/minute on /mcp because assistant platforms proxy many users through a small shared IP pool. Vehicle identity comes from NHTSA vPIC; all geometry is computed locally.