AutoScout24 Scraper: Car Prices, Price Drops & Market Stats avatar

AutoScout24 Scraper: Car Prices, Price Drops & Market Stats

Pricing

from $0.36 / 1,000 listings

Go to Apify Store
AutoScout24 Scraper: Car Prices, Price Drops & Market Stats

AutoScout24 Scraper: Car Prices, Price Drops & Market Stats

Export AutoScout24 car listings from 8 EU countries by make and model, past the 4,000-result limit: price, mileage, year, fuel, gearbox, power, price rating, dealer. Track new cars, price drops and removed ads; get median prices by model and year.

Pricing

from $0.36 / 1,000 listings

Rating

0.0

(0)

Developer

Datasmithlab

Datasmithlab

Maintained by Community

Actor stats

0

Bookmarked

2

Total users

1

Monthly active users

6 days ago

Last modified

Share

Export car listings from AutoScout24 by make and model in 8 EU countries (Germany, Austria, Italy, France, Netherlands, Belgium, Spain, Luxembourg). Every row comes with price, mileage, first registration, fuel, gearbox, power in kW and hp, engine size, consumption, CO₂, AutoScout24's price rating (top / good / fair price...), condition, city and the dealer.

AutoScout24 shows at most 4,000 cars per search (200 pages). Type volkswagen/golf and the Actor checks the count. If there are more, it reads the search in price bands, so you get the whole market instead of the first 4,000.

Two things you won't get from a plain scraper:

  • Monitoring. Schedule it and every later run returns only what changed: NEW cars, PRICE_DROP and PRICE_UP with the old price and the % change, and REMOVED ads. Unchanged cars are not returned and not charged.
  • Market summary by model and year. One row per country, make, model and first-registration year: listings, median and quartile price, median mileage, median hp, price per 1,000 km, dealer share and the share AutoScout24 rates as a good or top price. Ready for residual-value and pricing work.

No login, no browser. Phone numbers, contact-person names and private sellers' identities are not collected. Only search result pages are read; individual ad pages are never opened.

What you get

  • Price. price: 18990 (EUR), priceLabel (top-price, good-price, fair-price, somewhat-expensive...), previousPriceOnSite when AutoScout24 shows a crossed-out old price.
  • The car. make, model, modelVariant, version, title, mileageKm, firstRegistration (2019-06), year, condition (used, new, demo...), fuel, transmission, powerKw, powerHp, engineCcm, fuelConsumption, co2Emission, isDamaged.
  • Where. country, zip, city.
  • Seller. sellerType (Dealer / Private). For dealers: dealerName, dealerId, dealerStreet, dealerRating, dealerRatingCount, dealerUrl.
  • url of the ad, imageUrl, photoCount.
  • Reports in the run's key-value store: SEARCHES (one line per search and price band: cars available, pages read, status), MARKET_SUMMARY and HEALTH.
You wantSet
All BMW 3 Series in GermanyMake / model bmw/3er
A whole makeaudi
The same model in Austria and Italy tooCountries Germany, Austria, Italy
Diesel automatics from 2018, under 100,000 km, under €25,000Fuel Diesel, Gearbox Automatic, First registration from 2018, Mileage to 100000, Price to 25000
Only dealers (or only private sellers)Seller
A daily new-car / price-drop feedOnly new or changed listings on, schedule the Actor daily
Median prices per model yearMarket summary by model and year on
A search you already have openpaste it in AutoScout24 search URLs

The make/model is written as in AutoScout24 links (https://www.autoscout24.de/lst/mercedes-benz/c-klasse). Unknown makes or models are reported in SEARCHES and are not charged.

Output example

{
"recordType": "listing",
"id": "autoscout24:657ea487-75f9-46a8-a28e-51ac463d30fd",
"title": "BMW 3er d Touring",
"make": "BMW",
"model": "3er",
"modelVariant": "320",
"price": 10999,
"currency": "EUR",
"priceLabel": "top-price",
"mileageKm": 155932,
"firstRegistration": "2016-06",
"year": 2016,
"condition": "used",
"fuel": "Diesel",
"transmission": "Automatik",
"powerKw": 140,
"powerHp": 190,
"engineCcm": 1995,
"country": "DE",
"zip": "36043",
"city": "Fulda",
"sellerType": "Dealer",
"dealerName": "Autodeal24 / KFZ Vermittlungsagentur",
"url": "https://www.autoscout24.de/angebote/bmw-3er-...-657ea487-75f9-46a8-a28e-51ac463d30fd"
}

In monitoring mode rows also carry changeType, previousPrice, priceChange, priceChangePct and firstSeenAt.

Market summary row:

{
"recordType": "marketSummary",
"country": "AT", "make": "Volkswagen", "model": "Golf", "year": 2018,
"listings": 28, "medianPrice": 13995, "p25Price": 11990, "p75Price": 17567.5,
"minPrice": 7999, "maxPrice": 19990, "medianMileageKm": 154800, "medianPowerHp": 116,
"medianPricePer1000Km": 91.21, "dealerShare": 0.3929, "goodPriceShare": 0.3214, "priceDropOnSiteCount": 0
}

Pricing

Pay per event: a small start fee, then per listing row and per market summary row. Filtered-out cars and, in monitoring mode, unchanged cars are free. Set Max listings to cap the cost.

Good to know

  • Cars without a price ("price on request") are not included when a search is split into price bands. HEALTH.splitCoverage shows how much of the original count the bands covered.
  • REMOVED is only reported when every search was read completely in this run and the previous one, and never when more than half of the remembered cars vanish at once (that is far more likely a site change than a real sell-out).
  • Requests go out without a proxy. If AutoScout24 blocks them, the Actor retries through a residential proxy (at most 3 new IPs) and then stops and says so. It never tries to solve captchas.
  • Use the same settings for scheduled runs: monitoring memory is kept per set of searches and filters.

FAQ

Why not every field from the ad page (equipment list, description)? AutoScout24's robots.txt does not allow automated access to ad pages, so the Actor only reads search results. You still get the url of each ad.

Other countries' websites? All 8 markets are searched through autoscout24.de with AutoScout24's own country filter, so model names are the German ones (3er, c-klasse, golf).

Is this affiliated with AutoScout24? No. It reads public search pages at a polite pace and identifies itself honestly.