To scrape Immowelt without getting blocked, you need two things at once: a genuine browser engine and a non-datacenter IP address. A Python requests call with a Chrome User-Agent header will still be rejected, because the block happens before the server reads any HTTP header. Immowelt.de is one of Germany's leading real estate platforms, hosting thousands of residential, rental and commercial property listings whose prices and availability change frequently. Collecting that data manually cannot keep pace with a market that updates daily, which is why automated extraction matters. By the end of this post you will know the exact Scrape API parameters that work, what each Immowelt page costs in credits, and a copy-paste Python snippet you can run today.
What Actually Blocks Scrapers on Immowelt
The block happens below the HTTP layer. When your client opens a TLS connection, it sends a ClientHello message that lists its supported cipher suites, extensions and elliptic curves. As Cloudflare Bot Management documents, detection begins before HTTP headers are processed: bot systems use TLS fingerprinting of the ClientHello to classify clients. A Python requests call produces a fingerprint that is immediately recognisable as a non-browser client, regardless of the User-Agent string you set.
Setting a browser User-Agent on a requests call is like wearing a name badge that says "human" while walking through a fingerprint scanner. The scanner reads your hand, not the badge.
Even when a client passes the TLS check, a second layer fires: a JavaScript challenge that the server issues as an interstitial page. A headless browser running on a datacenter IP still fails here, because datacenter IP ranges are commonly classified and blocked by anti-bot systems. RFC 8446, which defines TLS 1.3 and its ClientHello structure, is the foundation that makes this fingerprinting possible at the protocol level.
The practical result: a datacenter IP combined with a non-browser TLS fingerprint is a near-certain block signal. You need to solve both problems simultaneously, which means a real browser engine and a residential or ISP-grade IP, both on the same request.
Takeaway: A user-agent header cannot fix a TLS fingerprint mismatch. You need js=true for the browser engine and proxy_type=premium for the IP reputation, both at once.
How to Scrape Immowelt Without Getting Blocked: The Working Setup
Think of a rotating proxy pool built from datacenter IPs as calling a phone number from a prefix that every spam filter already knows. The recipient sees the prefix and drops the call before you speak. Residential and ISP-grade proxies come from address ranges that belong to real internet service providers, so they pass the first IP classification check that datacenter ranges fail.
For sites with active bot defenses, the proxy tier is the single highest-leverage variable. JavaScript rendering with headless Chrome addresses the TLS fingerprint problem. The proxy tier addresses the IP reputation problem. You need both. On Immowelt, starting with the standard rotating tier and escalating only when you see blocks is a losing strategy: by the time you have confirmed the block, you have wasted time and added latency to your pipeline.
The right starting point is proxy_type=premium combined with js=true. Premium proxies route through residential or ISP-grade IPs. js=true launches headless Chrome, which produces a genuine browser TLS fingerprint and executes any JavaScript challenge the server issues.
Here is the minimum viable parameter set:
| Parameter | Value | Why |
|---|---|---|
proxy_type |
premium |
Routes through residential or ISP-grade IPs, not datacenter ranges |
js |
true |
Launches headless Chrome; produces a browser TLS fingerprint and executes JS challenges |
proxy_country |
DE |
Targets a German IP for a German-language site |
auto_captcha_solve |
true |
Attempts to solve supported CAPTCHA challenges if they appear |
block_resources |
true (default) |
Skips images and fonts; speeds up rendering without affecting listing data |
A request with proxy_type=premium and js=true costs 25 credits per successful request. Failed requests are not billed, so you only pay when you get real content back. The full parameter reference is in the Scrape API docs.
Takeaway: Start with proxy_type=premium and js=true. Add proxy_country=DE and auto_captcha_solve=true for production runs.
Try it yourself
Run this on your own pages
Get a free API key and make your first call in minutes. 5,000 free credits for 7 days, no card.
Scraping Immowelt Listings with the Recipe
The Scrape API supports a recipe parameter that applies a validated extraction configuration for a supported site and returns named fields as JSON, rather than raw HTML you have to parse yourself. The Immowelt recipe page shows the fields returned, a live sample response and the credit cost per call. The recipes catalogue lists every supported site.
A recipe call costs the same credits as the equivalent direct call with the recipe's settings: 25 credits for a premium-proxy JavaScript-rendered page. When you select a recipe in the Playground or Request Builder, the required settings are locked automatically, so you cannot accidentally under-configure the call.
The recipe returns structured listing data including property title, type (apartment, house, commercial, office), price, area, rooms, address and agent contact details from public listing pages. That is enough to feed a database, a dashboard, a CRM or an investment model directly, with no HTML parsing on your side. For the full field list and a live sample, see /recipes/immowelt.
Automation removes the manual effort, errors and operational costs that come with collecting listing data by hand. A site like Immowelt publishes listings across major German cities including Berlin, Munich, Hamburg, Frankfurt, Cologne, Stuttgart, Düsseldorf and Leipzig, and those listings change as properties are added, updated or removed. An automated pipeline keeps your dataset current without human intervention.
Takeaway: Use recipe=immowelt for listing URLs. The recipe returns structured JSON at 25 credits per page, ready for databases, dashboards, CRM systems and investment models without a parsing step.
Python Code Walkthrough
The snippet below fetches a Cologne apartment purchase search page using the recipe and prints the structured listing fields. It then shows how to check your account balance using the /account endpoint.
import requests
import json
API_KEY = "YOUR_API_KEY"
BASE_URL = "https://api.ujeebu.com/scrape"
# Fetch an Immowelt search results page with the recipe
params = {
"url": "https://www.immowelt.de/suche/koeln/wohnungen/kaufen",
"recipe": "immowelt",
"proxy_type": "premium",
"js": "true",
"proxy_country": "DE",
"auto_captcha_solve": "true",
}
response = requests.get(
BASE_URL,
params=params,
headers={"ApiKey": API_KEY},
)
if response.status_code == 200:
data = response.json()
listings = data.get("listings", [])
print(f"Listings returned: {len(listings)}")
print(json.dumps(listings[:2], indent=2, ensure_ascii=False))
else:
# A non-200 indicates the request did not complete successfully.
# Failed requests are not billed. Log and retry.
print(f"Request failed: {response.status_code}. No credits charged.")
# Check your remaining credit balance
account = requests.get(
"https://api.ujeebu.com/account",
headers={"ApiKey": API_KEY},
)
if account.status_code == 200:
info = account.json()
print(f"Credits used: {info['used']} of {info['quota']}")
A few things worth noting. Check response.status_code before parsing: a non-200 means the request did not complete successfully, and failed requests are not billed, so log it and retry rather than treating an empty result as valid data. The /account endpoint returns your current quota, used credits and next billing date, which is the reliable way to track consumption across a large run.
To scrape multiple cities, loop over a list of search URLs, one call per URL. Use proxy_session only when you need session continuity across paginated results for the same search; otherwise omit it and let the API rotate proxies freely. Implement your own retry policy: if a request returns a non-200, wait briefly and retry before moving to the next URL.
The data the recipe returns arrives as JSON, which means you can insert it directly into a database, export it as CSV or Excel for an investment model or dashboard, or pipe it into a CRM for lead generation. The gathering, filtering and structuring steps that manual collection requires are handled by the API call itself.
Takeaway: Check response.status_code before parsing. Use /account to monitor credit consumption. Implement retries client-side for non-200 responses.
What Immowelt Data Is Good For
Immowelt.de publishes real estate listings across Germany, and listing data changes as properties are added, updated or removed. That frequency is what makes automated extraction valuable: manual collection cannot keep pace, and the errors that come with copying data by hand compound quickly across large inventories.
The structured JSON the recipe returns is ready for several workflows without further transformation:
- Price monitoring and market trend analysis: track price-per-square-metre movements across cities and neighbourhoods to identify where the market is moving. Property categories (apartment, house, commercial, office) can be broken out separately for more granular analysis.
- Investment opportunity identification: compare asking prices against area averages to surface undervalued properties or districts with high rental yield potential.
- Lead generation: extract publicly listed agent contact details for targeted outreach. Real estate professionals use this to connect with agents or property owners whose listings match specific criteria.
- Real estate agency inventory management: agencies managing large property inventories use extraction to keep their own systems current without manual re-entry.
- Aggregator and comparison platforms: power a rental or purchase comparison site with structured listing data delivered as JSON, ready to insert into a database or render in a dashboard.
- Market research: study housing and rental trends across German cities for academic, journalistic or commercial research purposes. Market research firms use extraction to track demand patterns and price movements over time.
Because the recipe returns JSON directly, the data arrives in the shape your downstream system expects. Export it as CSV or Excel for spreadsheet-based investment models, load it into a CRM for sales workflows, or feed it into a dashboard for ongoing monitoring.
For ongoing data needs, a scheduled extraction pipeline keeps your dataset current automatically. If you need that pipeline built, hosted and maintained without running any infrastructure yourself, the custom scraping service is worth considering. The service also handles scraper maintenance when Immowelt updates its page structure, which removes the monitoring burden from your team.
Takeaway: The recipe's JSON output is database-ready and exports cleanly to CSV or Excel. No parsing step means less code to maintain when Immowelt updates its page structure.
Credit Cost and Pricing at Scale
A request with proxy_type=premium and js=true costs 25 credits. Here is what that means across each plan, using the plan credit allocations from the pricing page:
| Plan | Monthly credits | Pages at 25 credits each | Monthly cost |
|---|---|---|---|
| Starter | 300,000 | ~12,000 | $49 |
| Growth | 1,000,000 | tens of thousands | $119 |
| Pro | 3,000,000 | ~120,000 | $299 |
| Business | 15,000,000 | ~600,000 | $999 |
Failed requests are not billed, so your effective page count reflects successful fetches only. Annual billing saves 10% across all plans. The pricing page has the full breakdown.
The free trial includes 5,000 credits with no credit card required, which is enough for about 200 requests at 25 credits each when using premium proxy and JavaScript rendering. That is sufficient to validate the recipe output and test your parsing logic before committing to a plan.
Takeaway: Size your plan against your expected monthly page volume. At 25 credits per page, the Pro plan covers around 120,000 Immowelt pages a month.
When to Use the Custom Scraping Service Instead
The Scrape API with the Immowelt recipe is the right tool when you control the schedule, handle retries in your own code and need the flexibility to change cities, filters or property types on the fly. It is also the right tool when you are building a product that needs real-time or near-real-time data.
The custom scraping service makes more sense in two situations. First, when you need a scheduled data feed delivered to your database or storage bucket without running any infrastructure yourself. Second, when Immowelt changes its page structure and you need the scraper fixed without your involvement. The service builds, hosts, monitors and fixes the scraper for you, and delivers clean JSON, a scheduled export or a data feed. Custom projects start from $500, quoted per project.
A useful rule of thumb: if you find yourself writing retry logic, monitoring jobs and schema-migration code every time Immowelt updates its layout, the custom service is worth a conversation. If you want control and flexibility, the API is faster to start and cheaper at scale.
The Immowelt site guide has the full parameter reference, the recipe field list and a live request builder. The free trial gives you 5,000 credits with no credit card, enough to test about 200 pages at 25 credits each before you commit to a plan.
FAQ
Why does Immowelt block my scraper even when I set a browser user agent?
Because the block happens at the TLS layer, not the HTTP layer. As Cloudflare Bot Management documents, detection begins before HTTP headers are processed. A Python requests call produces a TLS fingerprint that is recognisable as a non-browser client regardless of the User-Agent string. You need js=true to launch headless Chrome and produce a genuine browser-grade TLS fingerprint.
Which proxy type do I need to scrape Immowelt reliably?
Use proxy_type=premium. Premium proxies route through residential or ISP-grade IP addresses that are not classified as datacenter ranges by anti-bot systems. The standard rotating tier uses datacenter IPs that are commonly flagged. Combine proxy_type=premium with js=true to address both the IP reputation problem and the TLS fingerprint problem in a single call.
How many credits does one Immowelt page cost with the recipe?
25 credits per successful request. That is the cost of a premium proxy with JavaScript rendering, which is what the Immowelt recipe uses. Failed and blocked requests are not billed.
Can I search Immowelt listings without supplying a URL?
Pass a full Immowelt search URL as the url parameter, for example https://www.immowelt.de/suche/koeln/wohnungen/kaufen, and use recipe=immowelt. The recipe page documents any search-mode parameters the recipe supports, including how to filter by city, property type and deal type.
Are blocked or failed Immowelt requests charged?
No. Credits are charged only on successful responses. If the request is blocked, times out or otherwise fails to complete, no credits are deducted.
How do I scrape multiple Immowelt cities at scale without getting banned?
Loop over your city list and call the API for each search URL. Use proxy_session only when you need session continuity across paginated results for the same search; otherwise omit it. Implement your own retry policy client-side: on a non-200 response, wait briefly and retry. For a scheduled feed across dozens of cities delivered to a database automatically, the custom scraping service removes the infrastructure burden. For more on proxy strategy and avoiding blocks at scale, see the proxy guide and the IP classification deep dive.
Try it yourself
Run this on your own pages
Get a free API key and make your first call in minutes. 5,000 free credits for 7 days, no card.