TurboTenant Rentals

Find US long-term rentals

TurboTenant Rentals lets users search public long-term rental listings across the United States by location, see them on an interactive map, refine results by rent, bedrooms, bathrooms, property type, and pet policy, page through additional results in a carousel, and open a single listing's detail view with photos, amenities, and a link to apply on TurboTenant.

What you can ask

Help me find rentals in my area
Example path →

MCP connector

Server URL: https://gpt-app.turbotenant.com/mcp
Login: None
Protocol: MCP 2025-11-25
Instructions: TurboTenant rental listing tools. Route requests by user intent:
Did the user point at a specific listing already shown (id, index, or "this one")? → .
Otherwise, is this the first rental search in the conversation? → .
Otherwise, did the user change the location vs. the previous rental search (different city, zip, neighborhood, or state — e.g. "show me Denver" then "now Brooklyn", or "Denver" then "what about Aurora")? → .
Otherwise (same location, any other filter change — rent, beds, baths, pets, property type, or "show me more") → .
The map establishes location. The carousel refines within the established location. A changed city, zip, neighborhood, or state starts a new location search.
A location is the same if it refers to the same city/zip/neighborhood and the user is narrowing, not relocating. "Denver" → "Denver under $2k" is the same location. "Denver" → "Brooklyn" or "Denver" → "Aurora" is a location change.
Search constraints all callers must know: every search needs a city/address query or structured address. For same-location refinements, carry that previous location anchor forward. Money filters are whole dollars, never cents. Use property type "coop" for co-op. A generic pet-friendly request sends only petFriendly=true; a specific pet request sends only petsAllowed. The two fields are mutually exclusive, and multiple petsAllowed values mean ANY/OR, not AND. Results contain only active public TurboTenant listings.

Tools

Reads data3

Skills

None in the store listings or on the server

MCP connector

Server URL

https://gpt-app.turbotenant.com/mcp

None

Protocol

MCP 2025-11-25

Instructions

TurboTenant rental listing tools. Route requests by user intent:

Did the user point at a specific listing already shown (id, index, or "this one")? → .

Otherwise, is this the first rental search in the conversation? → .

Otherwise, did the user change the location vs. the previous rental search (different city, zip, neighborhood, or state — e.g. "show me Denver" then "now Brooklyn", or "Denver" then "what about Aurora")? → .

Otherwise (same location, any other filter change — rent, beds, baths, pets, property type, or "show me more") → .

The map establishes location. The carousel refines within the established location. A changed city, zip, neighborhood, or state starts a new location search.

A location is the same if it refers to the same city/zip/neighborhood and the user is narrowing, not relocating. "Denver" → "Denver under $2k" is the same location. "Denver" → "Brooklyn" or "Denver" → "Aurora" is a location change.

Search constraints all callers must know: every search needs a city/address query or structured address. For same-location refinements, carry that previous location anchor forward. Money filters are whole dollars, never cents. Use property type "coop" for co-op. A generic pet-friendly request sends only petFriendly=true; a specific pet request sends only petsAllowed. The two fields are mutually exclusive, and multiple petsAllowed values mean ANY/OR, not AND. Results contain only active public TurboTenant listings.