How to Scrape Zillow Listings and Phone Numbers
Zillow prohibits automated scraping. Learn the authorized API path, privacy limits, compliant workflows, and how to document listing data safely.

Direct answer: Do not build a scraper for Zillow listings or phone numbers. Zillow’s Terms of Use, updated October 28, 2025, prohibit automated queries intended to obtain information from its services, including screen scraping, database scraping, spiders, robots, crawlers, and bypassing CAPTCHA or similar safeguards. The terms also restrict copying and redistributing listing information and images. If you need Zillow data, request access through an approved Zillow API agreement or use a separately licensed real-estate data provider.
This guide explains what the restriction means, how to choose an authorized data path, how to process data after you receive it lawfully, and how to capture visual references without turning screenshots into a data-extraction system. It does not provide instructions for evading CAPTCHA, rotating identities, defeating rate limits, harvesting phone numbers, or reproducing Zillow’s content.
1. What Zillow’s terms prohibit
Zillow’s Terms of Use say users may not “conduct automated queries (including screen and database scraping, spiders, robots, crawlers, bypassing ‘captcha’ or similar precautions, or any other automated activity with the purpose of obtaining information from the Services).” That is a contractual rule for use of Zillow’s services. It is not a universal statement about the law in every country or jurisdiction.
The same terms limit copying to activity performed without automated processes for personal or qualifying professional use. They also restrict reproducing listing images, listing information, or other service content, or making it available through another website, application, or service, unless Zillow explicitly permits it.
| Requested activity | How to treat it |
|---|---|
| Downloading listing pages automatically | Prohibited by the cited Zillow terms when done to obtain information. |
| Parsing HTML for addresses, prices, or agent details | Part of an automated extraction workflow; do not implement without written authorization. |
| Bypassing CAPTCHA, bot checks, login controls, or rate limits | Do not attempt. Zillow’s terms expressly include bypassing CAPTCHA or similar precautions. |
| Copying listing photos or descriptions into your own product | Requires permission and must follow the applicable license and attribution rules. |
| Using an approved feed or API | Follow the agreement, approved-site conditions, field restrictions, retention rules, and redistribution terms. |
2. Why phone-number scraping needs extra care
Zillow’s September 2025 Website Privacy Notice describes collection of contact information such as phone numbers and information supplied in communications. It also describes call recordings or transcripts where permitted by law. That statement describes Zillow’s own handling of information; it does not grant an outside party permission to collect, publish, retain, or contact people using those numbers.
Whether a phone number may be used for a call, text, lead-routing system, enrichment product, or marketing campaign depends on facts that are missing from the title: your jurisdiction, the recipient’s consent, the purpose, the channel, and whether the number belongs to a consumer, agent, brokerage, or business. Get advice for the jurisdictions where you operate and document the source, permission, and permitted purpose for every contact record.
3. The authorized Zillow API route
Zillow’s API terms describe an API for preapproved licensees to retrieve certain residential real-estate and mortgage data. The terms mention areas such as property details, postings, mortgage rates, home valuations, and information about real-estate professionals. Approval, available fields, pricing, retention, and redistribution rights depend on the agreement. Public documentation is not a promise of unrestricted or anonymous access.

Recommended approval checklist
- Describe your product, users, domains, and the exact fields you need.
- Explain whether data is displayed to end users, used internally, or combined with other sources.
- Document how you will store credentials, audit access, delete records, and honor corrections.
- Ask specifically about phone numbers: source, consent status, masking, retention, and permitted outreach.
- Confirm whether listing photos, descriptions, addresses, valuations, and agent data have separate licenses.
- Obtain written approval and keep the current API terms with your compliance records.
Use an adapter instead of scattering provider assumptions
After approval, isolate the provider-specific client behind an interface. The following example is runnable against a local or authorized endpoint that you configure; it intentionally does not invent a Zillow URL or claim that a public endpoint exists.
from dataclasses import dataclass
from typing import Any
import os
import requests
@dataclass
class Listing:
provider_id: str
address: str | None
price: int | None
phone: str | None
class AuthorizedProvider:
def __init__(self, endpoint: str, token: str):
self.endpoint = endpoint
self.token = token
def search(self, city: str) -> list[Listing]:
response = requests.get(
self.endpoint,
params={"city": city},
headers={"Authorization": f"Bearer {self.token}"},
timeout=30,
)
response.raise_for_status()
rows: list[dict[str, Any]] = response.json()["listings"]
return [
Listing(
provider_id=str(row["id"]),
address=row.get("address"),
price=row.get("price"),
phone=row.get("phone") if row.get("phone_permitted") else None,
)
for row in rows
]
provider = AuthorizedProvider(
endpoint=os.environ["AUTHORIZED_LISTINGS_ENDPOINT"],
token=os.environ["AUTHORIZED_LISTINGS_TOKEN"],
)
for listing in provider.search("Austin"):
print(listing)
The important controls are deliberate: credentials come from environment variables, the client uses a timeout, non-success responses fail loudly, and a phone number is retained only when the provider marks it as permitted. Replace the endpoint and response schema only with values supplied by your signed agreement.
4. cURL, Python, and Node.js patterns for an approved endpoint
These examples show the shape of a request without fabricating a Zillow URL. Set AUTHORIZED_LISTINGS_ENDPOINT to the endpoint documented in your approved contract.
curl --fail --silent --show-error \\
-G "$AUTHORIZED_LISTINGS_ENDPOINT" \\
-H "Authorization: Bearer $AUTHORIZED_LISTINGS_TOKEN" \\
--data-urlencode "city=Austin" \\
-o listings.json
import os
import requests
r = requests.get(
os.environ["AUTHORIZED_LISTINGS_ENDPOINT"],
params={"city": "Austin"},
headers={"Authorization": f"Bearer {os.environ['AUTHORIZED_LISTINGS_TOKEN']}"},
timeout=30,
)
r.raise_for_status()
with open("listings.json", "wb") as file:
file.write(r.content)
const endpoint = process.env.AUTHORIZED_LISTINGS_ENDPOINT;
const token = process.env.AUTHORIZED_LISTINGS_TOKEN;
const q = new URLSearchParams({ city: 'Austin' });
const res = await fetch(`${endpoint}?${q}`, {
headers: { Authorization: `Bearer ${token}` },
signal: AbortSignal.timeout(30000)
});
if (!res.ok) throw new Error(`${res.status} ${await res.text()}`);
await Bun.write('listings.json', await res.arrayBuffer());
Do not add concurrency, retries, pagination, or caching until your provider’s contract documents the relevant limits. A retry policy that is safe for one API can violate another provider’s quota or freshness rules.
5. Data governance after lawful receipt
- Minimize: Store only fields required for the approved use.
- Provenance: Keep provider, retrieval time, license version, and record ID.
- Access: Restrict phone fields to roles that need them and log reads.
- Retention: Apply the shortest period allowed by the agreement and applicable policy.
- Corrections: Provide a process for removing or correcting a listing or contact record.
- Redistribution: Treat exports, public pages, analytics, and partner feeds as separate use cases requiring review.
6. Screenshots for documentation, QA, and design review
A screenshot can document a page state for an authorized internal workflow, but it is not a license to extract or redistribute Zillow content. Capture only pages you are allowed to access, avoid collecting personal information unnecessarily, and do not use screenshots to defeat a control or build a listing database.

Or skip the browser setup
ScreenshotNeo provides a website screenshot API and MCP server. Use it only for pages you are authorized to capture. The request below demonstrates a neutral page; change the URL to your permitted target. Full options are in the ScreenshotNeo documentation.
curl -G "https://api.screenshotneo.com/v1/shot" -d access_key=YOUR_API_KEY --data-urlencode url=https://example.com -o shot.webp
import requests
r = requests.get("https://api.screenshotneo.com/v1/shot", params={"access_key": "YOUR_API_KEY", "url": "https://example.com"}, timeout=90)
r.raise_for_status()
open("shot.webp", "wb").write(r.content)
const q = new URLSearchParams({ access_key: 'YOUR_API_KEY', url: 'https://example.com' });
const res = await fetch(`https://api.screenshotneo.com/v1/shot?${q}`);
if (!res.ok) throw new Error(`${res.status} ${await res.text()}`);
require('fs').writeFileSync('shot.webp', Buffer.from(await res.arrayBuffer()));
ScreenshotNeo removes cookie and consent banners, newsletter popups, and chat widgets before capture. Bot checks, blank pages, timeouts, failed loads, and cache hits are not billed; each response identifies the result with X-Page-Verdict and X-Billed headers. It also offers full-page and element capture, device presets, custom viewports, dark mode, retina scale, PDF output, custom CSS and JavaScript, click and wait actions, request blocking, headers, cookies, user agents, authorization, timezone, geolocation, transparent backgrounds, resizing, chosen cache TTLs, signed links, asynchronous jobs, bulk capture, usage data, and an OpenAPI specification. Its MCP server exposes take_screenshot, get_page_info, and capture_pdf for Claude, Cursor, and other MCP clients.
Plans include 1,000 screenshots per month free with no card; paid plans start at $5 for 3,000 screenshots. Every feature is available on every plan. Use the service for permitted visual capture, not for automated collection of Zillow listing or phone data. Create a free ScreenshotNeo account.
7. Troubleshooting compliant integrations
| Symptom | Likely cause | Fix |
|---|---|---|
| Access denied or CAPTCHA | The service is protecting access or your use is not approved. | Stop automation. Contact Zillow or your licensed provider; never bypass the control. |
| 401 or 403 from an approved API | Expired token, wrong environment, or an unapproved site. | Verify credentials and approved domains with the provider. |
| Phone field is absent | The contract or record does not permit that field. | Do not infer or enrich it from scraped sources; ask the provider about permitted alternatives. |
| 429 rate limit | Request volume exceeds the documented quota. | Honor retry-after, reduce concurrency, and request a higher quota if allowed. |
| Duplicate listings | Multiple provider IDs or updates were merged incorrectly. | Use the provider’s stable ID and preserve update timestamps. |
| Screenshot shows a consent banner | The page requires an interaction or the target is outside your permitted workflow. | Confirm authorization and use documented capture options; do not automate consent circumvention. |
8. Performance, reliability, and cost
For an approved data API, measure latency, error rate, freshness, pagination cost, and quota consumption using the provider’s documented limits. Queue work, apply bounded timeouts, and make writes idempotent with the provider record ID. Keep raw responses only when the agreement permits them. Encrypt tokens and separate development credentials from production credentials.
For screenshots, full-page captures and PDF rendering generally require more browser work than a viewport image. Use a viewport capture when a thumbnail is enough, cache only when the page may be reused, and choose a cache TTL that matches your documentation need. ScreenshotNeo’s no-charge verdicts for bot checks, blank pages, timeouts, failed loads, and cache hits help avoid paying for unusable captures; inspect the response headers in your logs.
9. FAQ
Is scraping Zillow illegal?
The cited source establishes a Zillow contractual prohibition on automated queries and copying. Whether conduct also violates a law depends on jurisdiction, access method, authorization, and purpose. Do not describe the terms as a universal legal rule.
Can I scrape only public listings?
Public visibility does not override Zillow’s stated prohibition on automated queries intended to obtain information. Seek an approved feed or licensed source.
Can I collect agent phone numbers manually?
The terms and the applicable privacy, telemarketing, and data-protection rules still matter. Confirm permission, purpose, and redistribution rights before collecting or using any number.
Does a screenshot API make extraction permitted?
No. A screenshot tool changes how a page is rendered; it does not grant access or redistribution rights. Use it only for an authorized visual purpose.
Where should a startup begin?
Write a field-level data inventory, request approval from Zillow or choose a licensed provider, and make retention and contact permissions enforceable in code before importing records.


