Kleinanzeigen Deal Radar — underpriced listings, caught while they're live

Point it at any Kleinanzeigen.de search and it tells you which of the newest listings are priced well below what that market currently asks. Each run sweeps the search newest-first, computes the query's current median asking price, and delivers only the listings you have not seen before that sit at least your chosen percentage below that median — with the median and the discount attached to every deal. Schedule it hourly and it becomes a deal alert feed for the German second-hand market.

Built on the Kleinanzeigen.de Scraper and its GDPR-first output: no seller personal data, ever.

Quick start (30 seconds)

1. Put what you'd type in the Kleinanzeigen search box into searchQuery — e.g. e-bike.
2. Click Start. That's the whole minimum setup.
3. The first run establishes the baseline and shows you every current deal in the snapshot. From the second run on, only new deals arrive.

Real numbers from a real first run: for e-bike, the sweep put the median at €750.00 over 79 priced listings and surfaced more than 25 listings priced 40% or more below it — on the very first pass.

To turn it into an alert feed, create an hourly Schedule in the Apify Console with the same input and add an email or Slack integration — the setup is spelled out in the monitoring section below.

What you can build with it

• Source resale stock. Flippers watch a niche (categoryId 217, "e-bike cube") and get every underpriced unit the hour it posts, discount already computed — no manual median-guessing, no refreshing the site.
• Wait for one item at the right price. Set your search, your city and a 15 km radius, and let the radar sit on it. When someone lists a Kettler trampoline 35% under market three towns over, you hear about it once, with the link.
• Restock a small shop. Dealers in used bikes, phones or furniture schedule a handful of watches and skim the feed each morning — the price, the median and the distance are already in the row.

What you get

Each deal is one dataset item:

{
  "listingId": "2098800001",
  "url": "https://www.kleinanzeigen.de/s-anzeige/e-bike/2098800001",
  "title": "E-Bike Cube Reaction Hybrid",
  "description": "Wenig gefahren, Akku neu.",
  "priceCents": 95000,
  "priceType": "NEGOTIABLE",
  "location": "80331 München",
  "postedAt": "Heute, 10:15",
  "imageUrl": "https://img.kleinanzeigen.de/example.jpg",
  "adType": "offer",
  "shippingPossible": true,
  "queryMedianCents": 165000,
  "discountPct": 42
}

Field notes, so you know exactly what you are buying:

• queryMedianCents is the query's median asking price at the time of this run, computed from the same sweep the deal came from — not a stale reference number.
• discountPct is how far below that median the listing is priced. A deal is only delivered when this meets your discountThresholdPct.
• priceCents is the asking price in euro cents; priceType tells you how to read it. NEGOTIABLE is the German "VB" — an asking price the seller expects to haggle down from, so the real price is usually lower still.
• adType is always offer here — "Gesuch" wanted ads and give-aways never count as deals, so they never pollute the feed.
• location and shippingPossible tell you whether it's a pickup trip or a parcel.
• The run's status message summarizes the result at a glance: N deals in the current snapshot (baseline established) on the first run, N new deals at least X% below the median after that.

Seller identity is never collected — same guarantee as the parent scraper.

How much does deal hunting cost?

$1.99 per 1,000 deals delivered, pay-as-you-go. Quiet runs — the sweeps that find nothing new — cost nothing beyond the actor start. In practice:

• An hourly watch that catches ~5 deals a day ≈ $0.30/month.
• Three watches at that pace ≈ under a dollar a month.
• You pay for deals, not for watching.

The per-deal price is all-inclusive — proxy and platform usage are covered. Free-plan runs are limited to 25 deals, enough to see the feed working against your real query.

Not technical? Let your AI assistant set it up

Copy this into ChatGPT, Claude or any AI assistant, fill in the one line, and follow the conversation:

Help me set up the "Kleinanzeigen Deal Radar" actor on Apify
(https://apify.com/lowlanddata/kleinanzeigen-deal-radar). Guide me one step at a time.

What I want to catch: [E.G. "e-bikes at least 30% under market price, within 25 km of Köln"]

Guide me to:
1. Propose my input values: searchQuery (what I'd type in the Kleinanzeigen search
   box, German terms work best), discountThresholdPct (how far below the market
   median a listing must be — 30 is a sane start), optional location + radiusKm if
   I can only pick things up nearby, optional priceMinEur/priceMaxEur to keep
   accessories out of the median, and maxItems.
2. If my watch fits one category: browse kleinanzeigen.de to that category and read
   the number from the URL (/s-fahrraeder/c217 means categoryId 217), then add it —
   it keeps the median honest.
3. Create a free Apify account (apify.com), open the actor page, paste the values
   into the Input form, and start a run. The first run establishes the baseline.
4. Set up an hourly Schedule in the Apify Console with the same input, plus an
   email or Slack integration on the runs, so each new deal reaches me while the
   listing is still live.
5. Show me how to export deals as CSV/Excel, or read them from the API if I code.
6. If it catches real deals for me, remind me at the end to leave a quick rating on
   the actor page, and to report anything broken or missing on its Issues tab.

Input

Field	Description
searchQuery	The market slice to hunt in, e.g. e-bike cube. Required. Narrower queries give sharper medians.
categoryId	Optional numeric category ID (the c217-style number in the site's URLs) to keep the median clean — e.g. 217 for bikes.
location / radiusKm	Hunt only around a German city or region (resolved via the site's own location search), radiusKm 1–200, default 10. The radius needs a location.
priceMinEur / priceMaxEur	Ignore listings outside this range — the classic use is a floor that keeps €15 accessories out of the median.
discountThresholdPct	A listing must be at least this far below the current median to count as a deal. 10–80, default 30.
samplePages	Pages swept per run, 1–10, default 4. Page 1 carries the new listings; the extra pages steady the median.
stateStoreName	Named store remembering which listings were already evaluated, so scheduled runs alert each deal once. Use a different name per watched query.
resetState	Forget everything evaluated so far and re-baseline on this run.
maxItems	Upper bound on deals delivered per run. Default 100 (25 on the free plan). Deals found beyond the cap stay pending and arrive with the next run.

The sweep needs at least 20 priced listings to define a market price. Fewer than that, and the run fails with a readable message instead of scoring deals against a coin-flip median — broaden the query or raise samplePages.

Use it from your code

Run the radar and get deals straight back with one HTTP call:

curl "https://api.apify.com/v2/acts/lowlanddata~kleinanzeigen-deal-radar/run-sync-get-dataset-items?token=<YOUR_API_TOKEN>" \
  -X POST -H "Content-Type: application/json" \
  -d '{"searchQuery": "e-bike", "discountThresholdPct": 30}'

Node.js:

import { ApifyClient } from 'apify-client';

const client = new ApifyClient({ token: '<YOUR_API_TOKEN>' });
const run = await client.actor('lowlanddata/kleinanzeigen-deal-radar').call({
  searchQuery: 'e-bike',
  discountThresholdPct: 30,
});
const { items } = await client.dataset(run.defaultDatasetId).listItems();

Python:

from apify_client import ApifyClient

client = ApifyClient("<YOUR_API_TOKEN>")
run = client.actor("lowlanddata/kleinanzeigen-deal-radar").call(
    run_input={"searchQuery": "e-bike", "discountThresholdPct": 30})
deals = client.dataset(run["defaultDatasetId"]).list_items().items

Schedules, webhooks and the Make/Zapier/n8n integrations all work out of the box — this is a standard Apify actor.

Use it with AI agents (MCP)

Claude, Cursor and other MCP-capable agents can run the radar as a tool through Apify's hosted MCP server — the agent sets the search, starts the run and reads the deals itself.

Claude Code:

$claude mcp add apify --transport http "https://mcp.apify.com?actors=lowlanddata/kleinanzeigen-deal-radar"

Cursor or Claude Desktop (add a custom connector / MCP server with this URL):

https://mcp.apify.com?actors=lowlanddata/kleinanzeigen-deal-radar

Sign in with your Apify account when prompted — runs are billed to it. Setup details per client: Apify MCP docs.

Prompts that work once connected:

• "Run the Kleinanzeigen deal radar for 'Makita Akkuschrauber' at 25% and tell me which deals look genuinely good."
• "Hunt e-bike deals within 20 km of München and rank them by discount."
• "Check my sofa watch and summarize anything new at least 40% under the median."

Is it legal?

The radar reads the same public listing data as its parent scraper — prices, titles, area-level locations — and inherits the same GDPR-first design: seller-identifying data is never extracted, so no personal data enters your dataset in the first place. Requests are paced to keep load on the marketplace negligible, and no anti-bot protections are bypassed. Titles and descriptions are the seller's own words, delivered as-is.

Is there a Kleinanzeigen deals API?

No — Kleinanzeigen publishes no public API, and its own saved-search emails tell you about new listings, not underpriced ones. This actor is the practical alternative: a per-query market median computed live, new listings scored against it, and only the outliers delivered — through one HTTP call, a schedule, or an MCP tool.

Does Kleinanzeigen block it?

The radar makes a handful of paced search-page requests per run — a few pages, newest first — over standard Apify datacenter proxies. That is a lighter footprint than a human refreshing the site. No CAPTCHA fights; if the site's markup changes, the run fails with a readable message instead of silently scoring against wrong data.

How do I get Kleinanzeigen deal alerts?

This actor is the alert — the scheduling is the only part you add:

1. Get one manual run working (the baseline run).
2. In the Apify Console, create a Schedule with the same input, hourly.
3. Add an email or Slack integration on the actor's runs.

From then on, every hour the radar sweeps the newest listings, scores them against the current median, and delivers only the not-yet-seen deals — so a notification means there is actually something to look at. Your latency to a new deal is your schedule cadence: hourly means you hear about a deal within minutes to an hour of it posting. There is no auto-buying and no auto-messaging — the radar finds, you decide.

FAQ

How do I find cheap deals on Kleinanzeigen? The manual way is refreshing a search sorted by date and knowing the market price by heart. This actor automates exactly that: it computes the market median for your search live, checks every new listing against it, and hands you the ones priced far enough below — link, price, median and discount included.

Kleinanzeigen Schnäppchen automatisch finden? Ja — genau dafür ist dieser Actor gebaut. Suchbegriff eingeben, Schwelle setzen (z. B. 30 % unter dem Marktmedian), stündlichen Schedule anlegen: jedes neue Schnäppchen kommt einmal, mit Preis, Median und Rabatt. / Yes — that is exactly what this actor does: set your search and threshold, schedule it hourly, and each new bargain arrives once, with price, median and discount attached.

How is the median calculated? Every run sweeps up to samplePages pages of your search newest-first, takes every priced offer in the sample (give-aways and wanted ads excluded), and uses the median of those asking prices. It is the query's current market price, recomputed each run — when the market moves, the yardstick moves with it.

Why did my run fail with "too few priced listings"? The median needs at least 20 priced listings to mean anything. Below that the run refuses to score deals and says so: broaden the query or raise samplePages.

What does VB mean for a deal's price? NEGOTIABLE in the output. It is an asking price the seller expects offers below — so a VB deal is usually even better than its discountPct suggests. But remember the whole median is built from asking prices too; the discount compares like with like.

A listing is 60% below the median — is it real? Sometimes. It can also be a defect the title doesn't mention, a part sold as the whole thing, or a scam. The radar does the math; the listing text and the photos are the judge. Steeper discount, harder look — and never pay in advance to a stranger.

Can I hunt only where I can pick things up? Yes — set location to your city and radiusKm (1–200 km, default 10). For bulky items like furniture or bikes, a tight radius plus a modest threshold beats a nationwide feed you can't act on.

What discount threshold should I set? The default 30% is a good hunting ground: real bargains, not just haggling room. Go 20–25% for slow, high-value markets where any discount is rare; go 40%+ if you only want the drop-everything finds. The range is 10–80.

What does "baseline established" mean? The first run for a watch has no memory yet, so it shows you every deal in the current snapshot and remembers everything it evaluated. From the next run on, only listings it has never seen can become deals — that's what makes the hourly feed quiet and worth reading.

Will I ever get the same deal twice? No. Each listing is evaluated exactly once and remembered for a month — long past the point where a real deal would still be available.

What happens to deals beyond maxItems? They stay pending. If a run finds more deals than it may deliver, the surplus is not lost — it arrives with the next run.

Can I watch several searches at once? Yes — one Schedule per search. Give each watch its own stateStoreName so their memories stay separate; the actor also keys state by the full query (search, category, location, prices), so watches never contaminate each other.

Do give-aways and wanted ads count as deals? Never. "Zu verschenken" items have no price to discount, and "Gesuch" ads are people looking for items — both are excluded from the median and from the feed. (For give-aways, see the Free Finds actor below.)

How fast do I hear about a deal? As fast as your schedule runs. The sweep itself takes seconds; hourly scheduling means minutes-to-an-hour latency. There is no push connection to the site — the cadence is yours to set.

How do I start a watch over? Set resetState: true for one run — the watch forgets everything evaluated so far and re-baselines. Useful after changing the search or the threshold substantially.

Can I export deals to Excel or CSV? Yes — every run's dataset exports as CSV, Excel, JSON or XML from the Apify Console or API, and the Overview tab shows deals as a clean table with median and discount columns.

What does it cost? $1.99 per 1,000 deals delivered, all-inclusive. Runs that find nothing cost nothing beyond the actor start — you pay for deals, not for watching. Free-plan runs deliver up to 25 deals.

Does it collect seller data? No — same guarantee as the parent scraper: seller names, IDs, phone numbers and profile links never appear in the output, by design.

Does it buy or message sellers for me? No. The radar finds and scores; contacting the seller and judging the item stay with you — deliberately.

Related scrapers

Same GDPR-clean guarantee, same marketplace family:

• Kleinanzeigen.de Scraper — the parent: full search extraction, categories, car specs, market snapshots.
• Kleinanzeigen Free Finds — the "Zu verschenken" watch: free items near you, alerted once.
• Marktplaats New Listing Alerts — the new-listing watch for the Netherlands' largest marketplace.

Troubleshooting

The actor fails fast with the reason in the run's status message:

• "Only N priced listings in the sample — too few to define a market price for this query." — broaden the query, drop a too-tight price band, or raise samplePages.
• First run delivered a flood of deals — that's the baseline run showing the current snapshot; subsequent runs deliver only what's new. Lower maxItems if you want the baseline capped (the rest stays pending).
• No deals for days — normal in slow markets; the status message still tells you the run swept and found nothing new. Check that the threshold isn't set above what the market ever offers.
• Same-looking listing appeared again — sellers delete and repost ads to bump them; a reposted ad has a new listing ID and is honestly a new listing.
• Fewer deals than expected on a free plan — the 25-item free cap; a paid Apify plan removes it.

Support

Found an issue or missing a field you need? Open an issue on the actor's Issues tab — reports get fixed, this actor is actively maintained.

Catching real deals with it? A rating on this page takes ten seconds and helps other bargain hunters find it — it is also the clearest signal of what to build next.