Houski Property Data
Server Details
Property data for 220M+ homes in North America, Europe and South Korea: search, stats and values.
- Status
- Healthy
- Uptime
- 100.0% over the last 24 hours
- Last Tested
- Transport
- Streamable HTTP · MCP 2025-11-25
- URL
- ChatGPT
- –
- Claude
- –
TDQS
Scored across 12 tools
Most tools target distinct resources/actions: field_list vs field_detail, search vs properties, location vs geocoding, map, aggregate, predict. The main ambiguity is avm_endpoint, which has no description and could be confused with predict_endpoint or properties_endpoint valuation data.
Names are consistently snake_case and mostly follow a resource_action or action_endpoint style. Minor deviation: most tools use an _endpoint suffix, but field_detail, field_list, and search_api_documentation do not.
12 tools is well within the ideal range for a property-data API and each seems to earn its place across search, retrieval, aggregation, mapping, geocoding, prediction, and documentation. The count is not bloated or thin.
The surface covers discovery, retrieval, aggregation, geocoding, mapping, prediction, fields, auth, and docs, which is strong for a read-only property-data API. Minor gaps: avm_endpoint is undocumented, and there is no explicit single-property detail tool separate from properties_endpoint.
Available Tools
12 toolsaggregate_endpointAInspect
Get statistical aggregations for market analysis. Billed per aggregate value returned. Aggregations: count, sum, mean, median, mode, min, max, quantile-10, quantile-25, quantile-40, quantile-60, quantile-75, quantile-80, quantile-85, quantile-90, quantile-95, quantile-99. Text and boolean fields only support count and mode, and only fields field_detail marks aggregatable are accepted. count counts the properties that hold a value for the field, so a count per country measures how well a field is covered there. Ask for several statistics in one call with the aggregations array instead of one call each.
Inputs: api_key (required), city, field, exclude, +12 more.
auth_endpointAInspect
Check that an API key exists and is enabled. Does not bill per call. Returns api_key_authorized boolean. It does not check the key's maximum monthly spend, so a billed tool can still refuse a key that passes here.
Inputs: api_key (required).
avm_endpointDInspect
Inputs: api_key (required), property_id (required), overrides, price_quote.
field_detailAInspect
Describe fields by name: type, accepted filter values, its source column if it has one, filtered, aggregated or predicted, which expand set it belongs to, its price per value returned, and its full description. Does not bill per call. Coverage differs by country, and the server instructions say how to check it cheaply with location_endpoint.
Inputs: fields (required).
field_listAInspect
List the name of every public property data field, one per line. Does not bill per call. Pass filter to list only the names that contain a piece of text, for example score_, demographic_, assessment, permit or expand_listing. Follow with field_detail for each field's type, accepted filter values, aggregation support, expand set and price per value returned.
Inputs: filter.
geocoding_endpointAInspect
Find properties near coordinates (reverse geocoding). Billed per value returned. Required: latitude, longitude. Optional: radius (km), shape (circle/square). This endpoint has no country, province, city, community or address parameters. To narrow by place, put filters such as country_abbreviation_eq=CA or city_eq=Calgary in filters, spelled exactly as location_endpoint returns them, letter case included. For active or marketable properties, add has_expand_listings_eq=true to filters, which is far cheaper than expanding listings just to filter on their presence.
Inputs: api_key (required), latitude (required), longitude (required), page, +9 more.
location_endpointAInspect
Get the places Houski holds property data for, as a hierarchy of countries, provinces (or states or regions), cities and communities, with links and slugs. Billed per location returned, a small fraction of a cent each. This is the source of every country_abbreviation, province_abbreviation, city and community the other tools take. There is no paging and every matching location comes back, so scope every call: resolution=country and resolution=province are small, while resolution=city or community with no country_abbreviation and province_abbreviation returns every city of every country in one very large response. Naming a level requires naming its parents, and the endpoint returns a 400 saying so if you skip one. filters keeps only the places where at least one property matches, which is the cheapest way to check coverage: resolution=country with filters=["parcel_number_regex=."] lists only the countries that hold parcel numbers.
Inputs: api_key (required), city, flat, filters, +5 more.
map_endpointAInspect
Get map data: pins, clusters, or heatmaps. Billed per cluster or heatmap cell returned, or per value returned for pins. Scope it with all four bbox coordinates or a polygon: with neither, every property in every country is in scope. A heatmap needs all four bbox coordinates, and a heatmap scoped by a polygon alone is refused. Modes: cluster=true, heatmap=true. There is no cap on how many properties a request may match, so set abort_over yourself if you want a request to give up past a certain number. For maps of active inventory, add has_expand_listings_eq=true to filters so listing-less parcels don't pad results.
Inputs: api_key (required), page, select, cluster, +16 more.
predict_endpointAInspect
Predict property values over time. Billed per predicted value, which is one field on one date for one property. Required: property_id, start_date, fields. Find the property_id first with search_endpoint or properties_endpoint. Supports batch requests and parameter overrides for scenario testing.
Inputs: fields (required), api_key (required), start_date (required), property_id (required), +5 more.
properties_endpointAInspect
Retrieve property data (hundreds of fields) with filtering, sorting, and expansion. Billed per value returned: each selected field's price for each home that has a value for it, so a field with no value for a home is not billed. Always set select= to limit fields. For comparables, 'good' properties, or active inventory, add has_expand_listings_eq=true to filters: a fast indexed boolean filter that excludes the long tail of listing-less parcels (use has_expand_listings_rent_eq=true for rentals). Use expand=listings,permits,assessments only when you actually need the related records, not just to filter on their existence (has_expand_permits_eq=true and has_expand_assessments_eq=true cover that). Take place names and abbreviations from location_endpoint. When select names a field with a source column, pass exclude=inferred unless the user wants inferred values too or the call looks up one named property. Set price_quote=true first when the row count is unknown. See search_api_documentation for full filter syntax.
Inputs: api_key (required), city, page, expand, +17 more.
search_api_documentationAInspect
Access API documentation pages for filter syntax, field references, and endpoint details. Does not bill per call. Pass an unknown page to get the full list of pages back.
Inputs: page (required).
search_endpointAInspect
Find property_id using fuzzy address matching. Billed per value returned. Handles typos, abbreviations, partial addresses. Pass country_abbreviation to keep every match in one country. Use returned property_id with other endpoints for detailed data.
Inputs: query (required), api_key (required), select, exclude, +3 more.
Tool Schema Changelog
Recent tool additions, removals, and schema changes observed during successful MCP inspections.
Loading tool schema changelog…
Related MCP Connectors
U.S. real-estate data: property records, AVM value + rent estimates, sale/rental listings.
Live U.S. property data: 147M homes, ZIP market stats, daily listing changes, address lookup.
Property intelligence: 180M+ US parcels — lookup, search, owners, hazards, permits, deeds.
Live sale and rental listings from Zillow, Rightmove, Idealista and 15 more portals in one schema.
Related MCP Servers
- AlicenseNot gradedqualityBmaintenanceProvides access to comprehensive US property data, including automated valuations, tax history, comparable sales, and ownership details, enabling real estate analysis and market insights.90 npmMIT
- AlicenseAqualityBmaintenanceEnables searching Zillow for-sale, for-rent, and sold listings with rich filters and retrieving complete property details as structured JSON, with no Zillow account required.255 npm5MIT
- FlicenseNot gradedqualityDmaintenanceProvides free property and parcel intelligence using official US government data APIs, county assessor records, and live listings, with tools for hazard risk checks, affordability calculations, and parcel ranking.-
- AlicenseNot gradedqualityCmaintenanceAccess 17M+ geocoded French property transactions (DVF), 22M+ DPE energy ratings, and 20M+ building records via MCP or REST API. Search transactions, market stats, comparables, price trends, rental yield, flip detection, and more.MIT
Glama MCP Gateway
Add one secure layer between your agents and this server.