Best Zillow Scrapers in 2026: 8 Tools Compared

Zillow and six other real-estate sites, API by API

Attempts that returned the real page, out of five, on the October 30, 2026 run. Seconds are each API's average latency on zillow.com in the run's published results. Scrape.do and Context.dev are not profiled on this page; their rows are here because they also returned every real-estate attempt.

Five of the six profiled APIs returned zillow.com on every attempt. Zillow is not the hard site in this set: 14 of the 17 APIs in the run passed it five times out of five, and it ranks 93rd hardest of the 100 sites. Realtor.com and Apartments.com are the hard ones, 11th and 23rd hardest, and only five APIs passed Realtor.com every time.

Speed on Zillow favors String this run. String returned zillow.com in 1.2 seconds, the fastest of all 17 APIs, and Zyte was next of the six at 2.2. Nimble took 12.0 and Bright Data 14.8. Across the seven real-estate sites, String was faster than Scrapfly on six. The exception was Idealista, String's slowest site at 31.7 seconds against Scrapfly's 30.7.

String: every attempt on all seven real-estate sites

String is a web access API for developers and AI agents that fetches any URL and returns the page as Markdown, JSON or raw HTML, billed only when content comes back. Its Web Access API makes one HTTP call against any URL. Proxy rotation, anti-bot handling, captcha solving and JavaScript rendering happen behind the endpoint. A request for the benchmark's Zillow page looks like this:

curl https://request.usestring.ai/v1/fetch \
  -H "Authorization: Bearer $STRING_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"url": "https://www.zillow.com/new-york-ny/", "format": "markdown"}'

Add a jsonSchema field to get listing fields back as JSON instead of a page, as described in the structured extraction docs.

Pricing. Usage-based, as of October 2026. Growth costs $99 a month, all of it usage, and bills $0.16 per 1,000 fetches on a standard proxy, $0.90 on a premium proxy, $0.80 for a browser request on a standard proxy, and $3.25 with both. New accounts start with 20,000 free standard fetches, blocked requests cost nothing, and unused credit rolls over.

Benchmark success. 35 of 35 attempts on the seven real-estate sites, and 98.6% across all 100 sites, first of 17, in the October 30, 2026 run.

Latency. 1.2 seconds on zillow.com, the fastest of the 17 APIs. Its latency score across the run was 4.24 seconds, also the fastest.

Strengths. One of five APIs to return all four US listing portals on every attempt in the October 30, 2026 run. 100% against AWS WAF in that run, which protects Redfin and Realtor.com in the harness. A hosted MCP server gives an agent the same fetch without local setup. String also publishes two Zillow Actors on Apify, covered in the Apify section below.

Weaknesses. No prebuilt Zillow parser on the API, so you write the schema or the parsing. No SOC 2 report yet.

Scrapfly: open Zillow parsing code and 34 of 35

Pricing. On November 9, 2026, Pro was $100 a month for 1,000,000 credits, with overflow at $3.50 per 10,000. Per the pricing page, a datacenter request is 1 credit, JavaScript rendering or the Unblocker 5, and residential proxies 25, so a page costs $0.10 to $2.50 per 1,000 before add-ons; the billing docs carry the full matrix. The pricing page says upstream 4xx and 5xx responses and failed scrapes are free. Credits do not roll over.

Benchmark success. In the October 30, 2026 run, 34 of 35 on the real-estate sites, with four of five on Realtor.com, and 86.8% across all 100, fourth of 17. The run used Scrapfly's anti-bot bypass on the residential pool.

Latency. 4.5 seconds on zillow.com, and 69.0 seconds on Apartments.com, the slowest of the APIs that returned that site.

Strengths. Scrapfly publishes a Zillow scraping guide and an open Zillow scraper template in Python, a working start if you want parsing code rather than a raw page. Scraping, browser, crawler and extraction APIs share one key.

Weaknesses. A bill that moves with the proxy setting, and slower times than String on six of the seven real-estate sites.

Bright Data: an unblocker, a Zillow Scraper API and a Zillow dataset

Pricing. Checked November 9, 2026: no plan near $100. The Web Unlocker, the product the benchmark tested, costs $1.50 per 1,000 requests pay as you go, or $499 a month for 383,000, with 5,000 free requests a month. The Zillow Scraper API costs $1.50 per 1,000 records pay as you go, or $499 a month for 384,000, with 5,000 free records a month. Its product page says failed deliveries are not charged, while the Scraper API FAQ says records that fail because the input was incorrect are still billed. The Zillow dataset is priced at $250 per 100,000 records with a $250 minimum order. The Web Unlocker bills some premium domains at a higher rate shown only in the dashboard; zillow.com was not on that list on November 9, 2026.

Benchmark success. In the October 30, 2026 run, 33 of 35 on the real-estate sites through the Web Unlocker, with three of five on Idealista, and 75.4% across all 100, tenth of 17. The Zillow Scraper API and the dataset were not tested.

Latency. 14.8 seconds on zillow.com and 63.5 on Realtor.com. Its latency score of 27.88 seconds across the run was 15th of 17.

Strengths. Three ways to buy the same data: raw pages, parsed records, or a file. All four US portals on every attempt. A parsed record saves the parser you would otherwise write and maintain.

Weaknesses. Speed, and billing rules that differ between the product page and the docs. Ask for a sample before you buy the dataset. See our Bright Data alternatives page.

Nimble: a Zillow template, and zero on Apartments.com

Pricing. Checked November 9, 2026: no plan near $100. Nimble's pricing docs list Extract at $1.00 per 1,000 URLs, so $100 buys 100,000 successful requests, and the max driver at $2.50 per 1,000. Only successful requests are charged, credits expire at the end of each month, and the first plan above pay as you go is Startup at $2,500 a month billed annually. Nimble also runs a zillow_plp Extract Template, shown in a real-estate cookbook and billed at $3.00 per 1,000 runs. We did not test the template.

Benchmark success. In the October 30, 2026 run, 25 of 35 on the real-estate sites and 68.0% across all 100, 12th of 17. The run used automatic driver escalation.

Latency. 12.0 seconds on zillow.com, the second slowest of the six APIs here after Bright Data, and 37.2 seconds on Zoopla.

Strengths. A Zillow template that returns listing fields by ZIP code, and five of five on Zillow, Redfin, Zoopla and Rightmove. A hosted MCP server is available.

Weaknesses. Nothing returned on Apartments.com, two of five on Realtor.com, 51.0% against Akamai and 40.0% against AWS WAF. See our Nimble alternatives page.

Firecrawl: every real-estate attempt, pages ready for an LLM

Pricing. On November 9, 2026, Standard was $99 a month billed monthly for 100,000 credits, or $0.99 per 1,000 pages; the page defaults to the annual view at $83. A scrape that returns no document is free, but a page that responds with a 403 or 404 costs a credit. Enhanced proxy requests cost the same single credit. Credits roll over on Scale and Enterprise.

Benchmark success. In the October 30, 2026 run, 35 of 35 on the real-estate sites and 81.2% across all 100, sixth of 17.

Latency. 4.9 seconds on zillow.com, and a latency score of 8.87 seconds across the run, third of 17.

Strengths. Crawl, map and scrape endpoints built for LLM input, all four US portals on every attempt, 100% against PerimeterX, and an open-source core.

Weaknesses. A blocked page that returns a document still bills, and credits on Standard do not roll over. See our Firecrawl alternatives page.

Zyte: per-site pricing, zero on Apartments.com

Pricing. Checked November 9, 2026: Zyte prices each website in one of five tiers and assigns the tier itself. At the $100 monthly commitment, an HTTP request costs $0.10 to $0.95 per 1,000 and a browser request $0.75 to $12.00. The pricing page shows a named site's price only after you enter its URL, so we cannot state Zillow's tier. The pricing docs say: "We review tier assignments quarterly. Affected customers receive 2 weeks notice of any tier changes." Bans and rate limits are free, but a 404 from the target counts as successful and bills. Unused commitment is not refunded.

Benchmark success. In the October 30, 2026 run, 25 of 35 on the real-estate sites and 79.6% across all 100, seventh of 17. The run used Zyte's HTTP response mode. Its one Zillow miss was an HTTP 520, Zyte's code for a ban it could not get past in time.

Latency. 2.2 seconds on zillow.com, the fastest of the six after String, and 1.5 seconds on Rightmove.

Strengths. Speed on the sites it returns, and a real-estate data offer: build it yourself on Zyte API, or have Zyte deliver the data to spec at a quoted price.

Weaknesses. Nothing returned on Apartments.com, one of five on Realtor.com, and 70.0% against AWS WAF. No Zillow template. See our Zyte alternatives page.

Apify: prebuilt Zillow Actors, including two of ours

Apify's store sells scrapers called Actors, each built and priced by its own developer. We profile the Zillow Actor that Apify maintains and the two that String publishes. Disclosure: String earns from the second and third.

Pricing. As of November 2026 (checked November 9), Apify's plans are Free with $5 of monthly usage, Starter at $19 a month, Scale at $199 and Business at $999, and unused prepaid usage expires at the end of each billing cycle. The price per 1,000 results falls as the plan rises:

• Zillow Search Scraper, maintained by Apify: $2.30 on Free, $2.00 on Starter, $1.70 on Scale and $1.30 on Business.
• Zillow Property Listings Scraper, by String: $1.75 on Free, falling to $1.00 on Business.
• Zillow Property Detail Scraper, by String: $3.00 on Free, falling to $1.25 on Business.

Benchmark success. The Zillow Actors were not tested. In the October 30, 2026 run the benchmark ran Apify's general Web Fetch Actor, which returned zillow.com on five of five attempts, 34 of 35 real-estate attempts and 79.4% across all 100 sites. For a usage signal, Apify's Zillow Search Scraper reports 8.5K total users, 707 monthly active users and a 5.0 rating from 19 reviews on its store page. String's two Zillow Actors each show 2 total users.

Strengths. No parser to write. The Zillow Search Scraper returns properties for sale and rent from Zillow search. String's listings Actor returns about 41 listings per location, and its detail Actor adds lot size, year built and the listing description, up to 200 properties per run. Apify adds scheduling and storage.

Weaknesses. Each Actor is maintained by its own developer, so a Zillow redesign can break one until its maintainer fixes it. Two of the three Actors read paths that Zillow's robots.txt disallows for general crawlers: the Zillow Search Scraper needs search URLs that carry a searchQueryState parameter, and String's listings Actor reads /homes/ search pages. String's detail Actor reads /homedetails/ pages, which are not disallowed. The detail Actor's listed fields do not include photos, price history, tax history or school data.

Zillow: licensed APIs from Zillow Group

Pricing. Not published. The Zillow Group developer portal lists close to 20 APIs across listings, rentals, mortgages, public records and Zestimates, and each one is requested through an access form.

Benchmark success. Not tested. A licensed API is not a scraper, so the harness has nothing to call.

Strengths. The route Zillow Group names: it says it permits third parties to retrieve its data only through its API. The Zestimates API covers about 100 million US properties for commercial use. Zillow Group's Bridge data page offers public records for about 148 million properties across 3,200 counties.

Weaknesses. The MLS listings API is invite only, and access is at the discretion of each MLS partner. Bridge's page caps approved users at 1,000 calls a day per dataset and says you may not store the information locally; that page was last updated in 2021, so confirm the limits when you apply. None of these APIs reproduces the live search results a buyer sees on zillow.com.

What Zillow's terms and robots.txt say

Zillow's Terms of Use, updated October 28, 2025, list prohibited uses in Section 5. They include conducting "automated queries (including screen and database scraping, spiders, robots, crawlers" and bypassing captcha or similar precautions. Zillow Group's Bridge page adds: "We permit third parties to retrieve data from our site only through the API."

Zillow's robots.txt disallows /homes/ search paths for general crawlers, except a short list of exact index paths, along with any URL carrying searchQueryState and its internal /api/ and /graphql/ paths. It does not disallow /homedetails/ property pages in general, and it publishes sitemaps of property detail pages. The benchmark's test page, /new-york-ny/, is not on the disallow list.

What that means legally depends on how you access the site and what you do with the data. Our answer on whether web scraping is legal covers the US cases, the one on whether robots.txt legally prevents scraping explains why the file still matters as notice, and the hiQ v. LinkedIn ruling shows how a terms-of-use claim can outlast a win on the Computer Fraud and Abuse Act. None of this is legal advice. If Zillow data sits at the center of a product you sell, talk to counsel and look at the licensed APIs first.

The anti-bot cut

Real-estate sites sit behind several protection vendors. In the harness, Zillow is filed under PerimeterX, Redfin and Realtor.com under AWS WAF, Apartments.com under Akamai, Zoopla under Cloudflare and Idealista under DataDome. Success rates by protection vendor for the six APIs, across all 100 sites in the October 30, 2026 benchmark run. Site counts are in the column headers.

PerimeterX is the closest column: the six sit between 82.2% and 100%, and Firecrawl leads it. Zyte's 70.0% and Nimble's 40.0% on AWS WAF line up with their misses on Realtor.com. Nimble's 51.0% on Akamai lines up with its zero on Apartments.com. If your list reaches past Zillow, read the column for each site's protection before you read the leaderboard.

Does a Zillow scraper also work on Redfin, Realtor.com and Apartments.com?

Usually not all of them. Every page in this series carries one statistic computed from the run's per-site results, and this page's statistic is about the four US listing portals.

In the October 30, 2026 run of the Web Data Frontier Benchmark, 14 of the 17 APIs returned zillow.com on all five attempts. Follow those fourteen across the other three US portals:

• redfin.com: 11 of the 14 also returned it every time.
• apartments.com: 9 of the 14.
• realtor.com: 5 of the 14 (String, Scrape.do, Context.dev, Firecrawl and Bright Data).
• All four portals: the same 5 of the 14.

So a clean Zillow result predicts little about the rest of a US property dataset. If your dataset will grow past Zillow, the drop from fourteen APIs to five decides whether you need a second vendor, routing code and a second bill. Test the portal you will add next before you sign for the one you need today.

Why pricing is hard to compare across Zillow scrapers

1. The unit is different. String, Scrapfly, Firecrawl, Nimble and Zyte price per request or credit. Apify Actors and Bright Data's Zillow products price per result or record, so a search page of 40 listings is 40 units. Zillow's own APIs are licensed, with no public price.
2. A credit is not a page. One Firecrawl credit is one page at any proxy setting. One Scrapfly credit is a datacenter request, and the same page on residential proxies is 25.
3. The site sets the price. Zyte assigns each website a tier and shows it only after you enter the URL. Nimble's max driver costs two and a half times its standard Extract rate.
4. Failure billing differs. String, Scrapfly, Nimble and Zyte bill only successes, with an exception for target 404s at Zyte. Firecrawl bills a blocked page that returns a document.
5. Annual prices pose as monthly. On Firecrawl's pricing page as checked November 9, 2026, it shows $83 by default and bills $99 monthly. Nimble's first plan above pay as you go is billed annually.
6. Unused credit often expires. String's credit rolls over. Scrapfly's and Nimble's do not, and Zyte keeps an unused commitment.
7. A per-page price hides the parsing. A general API returns the page and you extract the listing fields. A prebuilt scraper returns the fields and prices each listing instead of each page.

What $100 a month buys

String's rates are as of October 2026, from its pricing page. The other prices come from each vendor's published pages, checked November 9, 2026.

• String, $99: $0.16 per 1,000 plain fetches, $0.90 premium, $0.80 browser, $3.25 browser on premium. Blocks free, credit rolls over.
• Scrapfly, $100: 1,000,000 credits, or $0.10 to $2.50 per 1,000 pages before rendering. No rollover.
• Firecrawl, $99: 100,000 credits, or $0.99 per 1,000 pages at any proxy setting.
• Zyte, $100 commitment: HTTP requests at $0.10 to $0.95 per 1,000, with Zillow's tier shown only after you enter the URL.
• Nimble: no $100 plan; $100 of pay as you go buys 100,000 Extract requests, or 40,000 on the max driver.
• Bright Data: no plan near $100; $1.50 per 1,000 Web Unlocker requests or Zillow records, pay as you go.
• Apify: no $100 plan; the $19 Starter plan plus usage puts the Zillow Search Scraper at $2.00 per 1,000 results.
• Zillow Group: no published price.

A Zillow search page carries many listings, so per-page and per-listing prices do not line up one to one. String's listings Actor returns about 41 listings from one location. Price your own mix of search pages and detail pages before you compare.

Alternatives to the Zillow API now that access is restricted

Zillow Group's MLS listings API is invite only, so most teams that ask for Zillow API access never get live listings from it. The other Zillow Group APIs are requested through a form, publish no price, and on Bridge cap approved users at 1,000 calls a day per dataset with no local storage. None of them returns the live search results a buyer sees on zillow.com.

Pick the replacement by what you used the API for. Prices are from vendor pages checked November 9, 2026, and String's are as of October 2026:

The first two rows are where most teams land. A scraping API costs least per listing at volume, because one search page holds about 40 listings for the price of one request. A prebuilt scraper costs more per listing but needs no parser. Both read zillow.com directly, which is what Zillow's terms prohibit, so a product built on Zillow data needs counsel's view and a look at the licensed route first.

Which Zillow scraper should you pick?

• You need Zillow plus Redfin, Realtor.com or Apartments.com. String or Firecrawl, two of the five APIs that returned all four on every attempt, or Bright Data if you also want parsed records.
• You need Zillow fast and at volume. String, at 1.2 seconds on zillow.com, after you test your own URLs.
• You want working Zillow parsing code to start from. Scrapfly's open template.
• You want parsed Zillow records without writing a parser. Bright Data's Zillow Scraper API, Nimble's Zillow template, or an Apify Actor after a test run to price it.
• You want a one-time bulk file. Bright Data's Zillow dataset, after you ask for a sample.
• You feed pages to a language model. Firecrawl, or String with Markdown output.
• Zillow data sits at the center of a product you sell. Zillow Group's licensed APIs, and counsel.

FAQ

How do I scrape Zillow listings?

Pick one of three routes. A scraping API such as String, Firecrawl, Scrapfly or Bright Data fetches a Zillow search or property page, and you parse the listings out. A prebuilt scraper such as an Apify Zillow Actor or Bright Data's Zillow Scraper API takes a city or URL and returns listing rows. A licensed API from Zillow Group returns data under contract. Five of the six APIs we profile returned zillow.com on five of five attempts on October 30, 2026. Read Zillow's Terms of Use first, because they prohibit automated queries.

Which scraping APIs handle PerimeterX and HUMAN challenges?

Across the nine PerimeterX sites in our 100-site benchmark, which include Zillow, in the October 30, 2026 run, Firecrawl passed 100% of attempts, String 97.8%, Scrapfly 93.3%, Zyte 91.1%, Bright Data 86.7% and Nimble 82.2%. Five of those six returned zillow.com on five of five attempts, so test the other PerimeterX sites on your list as well. Our answer on which scraping APIs handle PerimeterX covers all 17.

Does an API that scrapes Zillow also work on Redfin and Realtor.com?

Often not. Fourteen of 17 APIs returned zillow.com on every attempt in our October 30, 2026 run. Of those fourteen, 11 also returned redfin.com every time, 5 returned realtor.com, and the same 5 returned all four US portals including apartments.com: String, Scrape.do, Context.dev, Firecrawl and Bright Data.

Where can I buy a Zillow dataset instead of scraping it myself?

As of November 9, 2026, Bright Data sells Zillow datasets at $250 per 100,000 records with a $250 minimum order, and its page lists two datasets, property listings and price history. Zillow Group licenses data directly through close to 20 APIs, including Zestimates for about 100 million US properties, with access by request and no published prices. Zillow also offers downloadable regional data, such as its home value index, rent index and for-sale inventory, on its research data page for your own analyses.

Apify vs building my own scraper: which is cheaper long term?

As of November 9, 2026, an Apify Actor costs $1.00 to $3.00 per 1,000 Zillow results and needs no parser, so it wins while your volume is small. An API plus your own parser costs less per listing at volume, because one search page carries about 40 listings for the price of one request, and String bills $0.16 per 1,000 standard fetches on its $99 plan (as of October 2026). The parser is the cost you take on, and it breaks when Zillow changes its page.

Is scraping against a website's terms of service enforceable?

It can be. Zillow's Terms of Use prohibit automated queries and bypassing captchas, and a breach-of-contract claim is the risk that survived in hiQ v. LinkedIn, where LinkedIn won on contract after the Ninth Circuit sided with hiQ on the Computer Fraud and Abuse Act. Our legal answer covers how US courts have ruled. It is not legal advice.

How do I measure a scraping provider's real success rate before buying?

Run your own Zillow URLs, five attempts each, and count a request as passed only when the response contains text from the real page, such as the page title. A captcha page with HTTP 200 is a failure. Our benchmark harness does exactly this, with the Zillow URL and marker in its test file, and it runs with your own API keys.

Do I need a headless browser to scrape Zillow?

Not for the search page we tested. Scrapfly returned it on every attempt with its adapter asking for raw HTML without rendering, and Zyte returned four of five in its HTTP response mode. A browser still helps when you must click, scroll a map or load more results, and it adds cost and time.

Sources

Checked November 9, 2026 unless noted.

1. String, Web Data Frontier Benchmark: /benchmark and the October 2026 results post
2. Benchmark harness repository, the October 30, 2026 run file, the test file at the run's commit and Firecrawl's adapter pull request
3. String pricing, structured extraction docs and MCP docs
4. Scrapfly pricing, billing docs, Zillow guide and Zillow scraper template
5. Bright Data Web Unlocker pricing, Web Unlocker features, Zillow Scraper API, Scraper API FAQ and Zillow dataset
6. Nimble pricing, pricing docs, real-estate cookbook and MCP server
7. Firecrawl pricing, billing docs and open-source repository
8. Zyte pricing, pricing docs, error handling docs and real-estate data offer
9. Apify pricing, Zillow Search Scraper, String's Zillow Property Listings Scraper and Zillow Property Detail Scraper
10. Zillow: Terms of Use, robots.txt, benchmark test page and research data
11. Zillow Group: developer portal, Zestimates API, MLS listings API and Bridge data page