Compass Real Estate Scraper — Homes for Sale & Rentals
Pricing
from $4.18 / 1,000 compass listings
Compass Real Estate Scraper — Homes for Sale & Rentals
Scrape Compass.com real estate listings by city or URL: price, beds, baths, sq ft, address, ZIP and photos. About USD 5.22 per 1,000 listings.
Pricing
from $4.18 / 1,000 compass listings
Rating
0.0
(0)
Developer
Muhammad Afzal
Maintained by CommunityActor stats
0
Bookmarked
6
Total users
3
Monthly active users
9 days ago
Last modified
Categories
Features
Share
Compass Real Estate Scraper extracts public Compass.com listings by city or URL and returns price, beds, baths, square feet, street address, ZIP code, neighborhood, MLS number, listing agent, photo URLs and the listing link, one row per property. It is built for US property researchers, investors, proptech teams and agents who need Compass listing data in a spreadsheet or API, and it costs about USD 5.22 per 1,000 listings on the Free tier, with no subscription.
You can run it from Apify Console, the Apify API, a schedule, or an AI assistant. Each listing is read from its own public Compass listing page and checked against the dataset schema before it is stored and charged.
What data does the Compass real estate scraper return?
Every dataset row is one unique Compass listing, built from the data Compass publishes on that listing's page.
| Field | What it contains |
|---|---|
listingId, listingUrl | Compass listing ID and public listing URL |
price, priceText, currency | Asking price (monthly rent for rentals, closing price for sold homes) as a number, the displayed text such as $899,000, and USD |
bedrooms, bathrooms | Bed and bath counts; half baths count as 0.5 |
livingAreaSqFt, lotSizeSqFt | Living area and lot size in square feet |
address, city, state, postalCode, neighborhood | Street address with unit, city, two-letter state, ZIP code and the neighborhood Compass assigns |
latitude, longitude | Coordinates from the listing page |
listingType | for-sale, for-rent or sold |
status | Compass's status label, such as New, Active, Price Changed, Under Agreement, Coming Soon or Sold |
propertyType, description, mlsId, listedAt | Property type, listing description, MLS number and the date the listing went on the market (YYYY-MM-DD) |
agentName, agentPhone, brokerage | Listing agent or team, phone and brokerage shown on the listing page |
imageUrls | Up to 20 Compass photo URLs |
sourceUrl, searchUrl, scrapedAt, warnings | Listing page fetched, the city search it was matched to, UTC timestamp, extraction notes |
Fields Compass does not publish for a listing are null, never guessed.
The OUTPUT record in the run's key-value store summarizes the run: status (SUCCEEDED, EMPTY, BLOCKED or REJECTED), recordsCollected, chargedEvents, listing pages checked and skipped (by reason), index files read, the access route used, and warnings.
How do I scrape Compass listings by city or URL?
Give the Actor a city in "City, ST" form, a Compass city search URL, or Compass listing URLs. For a city, it finds that city's listings in Compass's public listing index (the sitemap Compass publishes for search engines), newest first, opens each listing page, and keeps the ones that match listingType. "New York, NY" covers the five New York City boroughs.
| Input | Default | Bounds and notes |
|---|---|---|
locations | ["Boston, MA"] (prefill) | US cities in "City, ST" form. With only startUrls, no city search runs |
startUrls | [] | Compass listing URLs (/homedetails/...) or city and neighborhood search URLs such as /homes-for-sale/austin-tx/, /for-rent/miami-fl/, /homes-for-sale/boston-ma/sold/. A search URL keeps its own listing type; price, bedroom and other filters inside the URL are not applied |
listingType | for-sale | for-sale (active, coming-soon and under-contract homes), for-rent (active rentals) or sold (closed sales); applies to locations only |
maxResults | 1 | 1–1,000 unique listings stored and billed per run, split evenly across cities |
Search for rentals in two cities:
{"locations": ["Miami, FL", "Austin, TX"],"listingType": "for-rent","maxResults": 100}
Use Compass URLs only (the Boston default runs only when neither locations nor startUrls is given):
{"locations": [],"startUrls": [{ "url": "https://www.compass.com/homes-for-sale/manhattan-ny/" },{ "url": "https://www.compass.com/homedetails/5-Dow-Rd-Boston-MA-02132/1Y7MSH_pid/" }],"maxResults": 50}
The defaults (one Boston listing) are a quick health check. Raise maxResults for real work. URLs outside compass.com are ignored with a warning.
How fast is it, and how many listings fit in one run?
Since 2026-11-06 Compass shows a browser check on its listing pages, so the Actor reads each listing page fully rendered, up to six at a time; one rendered page takes about 15 seconds. For-sale searches deliver about 12 to 15 listings per minute (50 Boston listings took 4 minutes in a 2026-11-06 test), because each listing page names the newest active homes for sale in its ZIP code and the Actor opens those next. Rental searches deliver about 3 to 6 listings per minute (5 Chicago rentals took 108 seconds in a 2026-11-07 test): most of a city's listing pages are homes for sale, so the Actor checks unit-numbered addresses first, where most rentals are. Sold searches start from Compass's recently-sold pages, which list each region's latest closed sales, so they also deliver about 3 to 6 listings per minute, newest sales first. In Alaska, Arkansas, Iowa, Montana, Nebraska, North Dakota, Oklahoma and South Dakota, which have no recently-sold page, and once a city's recent sales are used up, sold searches fall back to Compass's listing index, which mixes sold homes with closed rentals, and slow to about one listing per minute or less.
The default run timeout of 300 seconds fits about 50 for-sale listings. For larger runs, raise the timeout in run options: 1,000 for-sale listings take about 75 minutes. Neighborhood search URLs, such as /homes-for-sale/back-bay-boston-ma/, check the city's listings until they find ones in the neighborhood, so they are slower again; a sold neighborhood search took almost four minutes for one listing in a 2026-11-07 test, so raise the timeout for those. Each run has a page limit of about six listing pages per requested listing (twelve for rentals, eight for sold homes, plus twenty for a neighborhood); a run that reaches it stops, keeps the listings it has, and says so in its warnings.
If Compass lifts the browser check, the Actor reads listing pages directly again, one every 1.5 seconds, the pace Compass accepted for 400 pages in a row on 2026-10-31.
What does the output look like?
Two records from a Boston for-sale run (truncated; long fields shortened):
[{"listingId": "1Y7MSH_pid","listingUrl": "https://www.compass.com/homedetails/5-Dow-Rd-Boston-MA-02132/1Y7MSH_pid/","status": "New","listingType": "for-sale","price": 899000,"priceText": "$899,000","currency": "USD","bedrooms": 5,"bathrooms": 2.5,"livingAreaSqFt": 2652,"lotSizeSqFt": 7405,"propertyType": "Single Family","address": "5 Dow Road","city": "Boston","state": "MA","postalCode": "02132","neighborhood": "West Roxbury","latitude": 42.2809706,"longitude": -71.1668762,"mlsId": "73596308","listedAt": "2026-10-29","brokerage": "Coldwell Banker Realty","imageUrls": ["https://www.compass.com/m/1f6497b06fa1c91ac509f283cf29eda8e24059bc3c6d2cabab864e9768f6a825/origin.jpg"],"searchUrl": "https://www.compass.com/homes-for-sale/boston-ma/","scrapedAt": "2026-10-31T05:32:36.912Z","warnings": []},{"listingId": "1YOGZN_pid","listingUrl": "https://www.compass.com/homedetails/470-Massachusetts-Ave-Unit-3-Boston-MA-02118/1YOGZN_pid/","status": "Active","listingType": "for-rent","price": 3600,"priceText": "$3,600","bedrooms": 2,"bathrooms": 1,"livingAreaSqFt": 860,"address": "470 Massachusetts Avenue, Unit 3","city": "Boston","state": "MA","postalCode": "02118","neighborhood": "South End","mlsId": "73578250","searchUrl": null,"warnings": []}]
Both records come from 2026-10-31 runs; the second was requested by its listing URL, so searchUrl is null.
When should I use Compass Real Estate Scraper, and when not?
Use it when you need structured Compass listing data: price and size comparisons by city, rental supply snapshots, new-listing monitoring on a schedule, sold-price comparisons, or property records for a model, dashboard or AI workflow.
Choose a different tool when:
- You need agent emails and phone numbers for outreach: Realtor Leads Scraper or Realtor Agents Scraper.
- You need recorded sale prices for New York City: NYC Property Sales Scraper.
- You need owner, tax or assessor data: County Property Records API.
- You want listings from RE/MAX instead of Compass: RE/MAX Real Estate Listings Scraper.
- You need Canadian listings: Realtor.ca Scraper.
- You need Compass's search filters (price, beds, map area) applied exactly: the Actor matches by city and listing type only, so filter the dataset after the run.
It does not log in, read private MLS fields, contact agents, or solve CAPTCHAs.
How do I use it from an AI assistant?
Connect Claude, ChatGPT, Cursor or any MCP client to Apify's MCP server with this single-tool URL: https://mcp.apify.com/?tools=muhammadafzal/compass-scraper. The assistant can then run the scraper and read the results directly.
Example prompt: "Use Compass Real Estate Scraper to get 50 homes for sale in Austin, TX and rank them by price per square foot."
How do I run it with the Apify API?
JavaScript:
import { ApifyClient } from 'apify-client';const client = new ApifyClient({ token: process.env.APIFY_TOKEN });const run = await client.actor('muhammadafzal/compass-scraper').call({locations: ['Boston, MA'],listingType: 'for-sale',maxResults: 25,});const { items } = await client.dataset(run.defaultDatasetId).listItems();console.log(items);
Python:
from apify_client import ApifyClientclient = ApifyClient("MY_APIFY_TOKEN")run = client.actor("muhammadafzal/compass-scraper").call(run_input={"startUrls": [{"url": "https://www.compass.com/for-rent/miami-fl/"}],"locations": [],"maxResults": 25,})items = client.dataset(run.default_dataset_id).list_items().items
Results export from Apify Console as JSON, CSV, Excel, XML or RSS.
How much does it cost to scrape Compass listings? Pricing per tier
You pay per listing stored, plus a small start fee charged per started GB of allocated memory. Apify platform usage is included in these prices.
| Event | Free | Bronze | Silver | Gold, Platinum, Diamond |
|---|---|---|---|---|
| Compass listing (primary) | USD 0.005219 | USD 0.005088 | USD 0.004958 | USD 0.004175 |
| Actor start, per started GB of memory | USD 0.006251 | USD 0.006094 | USD 0.005938 | USD 0.005 |
Worked totals on the Free tier at the default 256 MB memory (one start event):
- 100 listings: 100 × 0.005219 + 0.006251 = about USD 0.53.
- 1,000 listings, the per-run maximum: about USD 5.23.
A listing is charged only after it passes schema validation and is stored. Listing pages that turn out to be another listing type, outside the city, or no longer available are skipped and not charged. The status message and the estimatedRecordCostUsd value in OUTPUT use the live per-listing price of your pricing tier.
Why did my Compass run return zero listings?
The usual reasons are a city name the Actor could not read (write it as "City, ST"), a city with no current Compass listings of that type, or Compass refusing requests. A run with no matching listings ends as EMPTY with a warning and no listing charges. If Compass refuses access before any listing page can be read, the run ends BLOCKED, is marked failed, and charges no listings; it never invents records. Retrying a few minutes later usually helps.
Compass protects its search pages and, since November 2026, its listing pages with a browser check. The Actor does not use the search pages: it reads Compass's public listing index directly and each public listing page through a rendering route managed by the Actor. Proxy settings are not part of the input.
FAQ
Can I use this as a Compass listings API?
Yes. Call the Actor through the Apify API or client libraries with a city or Compass URL, and read the listings back as JSON rows from the run's dataset.
Can I scrape Compass rentals and sold homes?
Yes. Set listingType to for-rent or sold for city searches, or paste a Compass rentals or sold-homes URL into startUrls. Sold records carry status: "Sold" and the closing price in price.
Does it return agent phone numbers?
It returns the listing agent's name, phone and brokerage when the listing page shows them. Use a dedicated agent-lead Actor for outreach lists.
Does it apply the filters in my Compass search URL?
No. A search URL sets the city (or neighborhood) and listing type; price, bedroom, map and other filters in the URL are ignored with a warning. Filter the dataset after the run.
How many Compass listings can I get per run?
Up to 1,000 listings per run, split evenly across the cities you list. Raise the run timeout for runs above about 50 for-sale listings.
Can I export Compass listings to Excel or Google Sheets?
Yes. Export the dataset as CSV or Excel, or connect Google Sheets, Zapier, Make or a webhook through Apify integrations.
Is it legal to scrape Compass?
The Actor reads only public pages and does not bypass logins. Compass's terms, MLS rules and privacy law still apply to how you collect and reuse the data, so review them for your use case.
Related Actors
- Zameen.com Scraper for Pakistan property listings
- Lokaleportalen.dk Scraper for Danish commercial property
- YATCO Scraper for yachts for sale
- JamesEdition Real Estate Scraper for luxury listings worldwide
- RE/MAX Real Estate Listings Scraper
- Realtor Leads Scraper
Legal and responsible use
This Actor collects publicly visible Compass.com listing data only. You are responsible for complying with Compass's terms, MLS and data-feed restrictions, copyright in photos and descriptions, and privacy law when you store or contact anyone from the results. Do not republish listings as a competing portal. The Actor is independent and not affiliated with Compass.
Last verified: 2026-11-07