About Mireye
Mireye is an API and Model Context Protocol server that gives AI agents cited, structured data about physical locations. An agent can ask for facts about a place, such as elevation, flood zone, slope, solar irradiance, nearest transmission line or parcel ownership, and receive each value with its source, fetch timestamp and confidence level. Sources include federal datasets such as USGS, FEMA, NOAA, USDA, EPA, the Census Bureau and Sentinel 2 imagery. The vendor is backed by Y Combinator.
It is aimed at developers building agents for work like data center siting, renewable energy siting, real estate and land diligence. The API offers geocoding, a catalog of 327 typed fields, natural language questions about US locations, parcel lookup with boundary and owner of record, and proximity calculations such as drive time, nearest airport or substation, and labor shed population. If a field is missing from the catalog, customers can request it, and a genuine gap becomes a queued build available to every caller.
Pricing is credit based, and every call costs a fixed number of credits. Free is $0 with 5,000 credits per month and 20 requests per minute. Build is $19 per month with 25,000 credits and 60 requests per minute. Growth is $99 per month with 120,000 credits, 300 requests per minute and a shared Slack channel. Enterprise is custom priced with custom SLAs. No plan bills overage, so usage stops at the allowance. Examples of credit costs include 1 credit per geocode, 10 credits per natural language question and 300 credits per parcel lookup.