Collecting Cars Scraper

Turn Collecting Cars into a structured auction dataset: every live lot with its current bid, and the full archive of ~33,000 finished lots with the price each one sold for. Scrape cars, motorbikes, watches and number plates with make, model, generation, year, mileage, bid count, reserve state, closing time, country and seller type, then optionally add VIN, exterior and interior colour, engine size, original market, the seller's write-up and every photo. Filter by country, make, model, year and price. JSON, CSV or Excel out.

💰 $3 per 1,000 lots. No monthly fee, no proxy bill, no subscription.

How it works

✨ Why use this scraper?

The collectingcars.com listing grid only shows the ~300 auctions that are live right now. Once a lot finishes it drops off the grid, and with it the result it made. This Actor reads the site's own search index instead, which still holds every finished lot, so the whole sold archive comes back with its prices.

• ~33,000 sold lots with their sale price. 99.9% of the archive carries one, with the final bid, bid count and date sold.
• Live and upcoming lots too: current bid, bid count, reserve state and the exact closing time, so you can follow an auction as it runs.
• No result ceiling. 250 lots per request with no cut-off, so the entire catalogue is reachable in about 140 requests.
• Filter the way a buyer does: country, make, model, year, price, drive side, seller type, no-reserve only and sale format.
• Optional detail pass adds VIN, engine size, exterior and interior colour, body style, registration, original market, the seller's full write-up, every photo, and view and watcher counts.
• Cars, bikes, watches, plates and parts. Collecting Cars and its sister site Watch Collecting share one catalogue, and both are covered.

🎯 Use cases

Team	What they build
Valuation & price guides	Sale prices by make, model, generation, year and country from ~33,000 real results
Dealers & traders	Sourcing lists of live lots in chosen countries, and comps for what a model actually fetched
Market analysts	Price trends over time, and how UK results compare with German, Swedish or Australian ones
Insurers & lenders	Evidence-based agreed values for collector-car policies and finance
Auction watchers	Daily feeds of new consignments with listedWithinHours, plus live bid tracking
ML / data teams	Training data that pairs spec, mileage, location and bidding activity with a real sale price

📥 Supported inputs

You don't need a URL. Set the filters and run. startUrls is optional:

Input	Behaviour
Nothing at all	Returns the auctions live right now
https://collectingcars.com/buy?refinementList[productMake][0]=Porsche	Search: the page's filters are translated into the same search the site runs
https://collectingcars.com/makes/Honda/48/NSX/18435	Make/model page: read as a make and model filter
https://collectingcars.com/for-sale/1990-honda-nsx-1	Single lot: scrapes just that lot, always with its full detail
https://watchcollecting.com/for-sale/...	Watch lot from the sister site

Paste as many URLs as you like. Results are de-duplicated by lot ID across all of them, so overlapping searches never bill twice, and maxItems is a total across every URL.

If the Actor cannot read any filter from a search URL you paste, it stops with an error rather than running. An unreadable URL would otherwise fall through to scraping the whole catalogue instead of your selection. A run given only single-lot URLs scrapes exactly those lots and nothing else.

Not supported: hosts other than collectingcars.com and watchcollecting.com, and the price-range sliders on the site's search page (set priceMin/priceMax on the input instead).

⚠️ Proxy requirement

None for the search itself, which is why a run without the detail option is fast and cheap. The lot pages behind the detail option sit behind Cloudflare and need a residential IP. One is built in, so leave proxy empty unless you want to use your own.

🔄 How it works

1. Your filters, or the URLs you pasted, become one search against the site's search index.
2. Lots come back 250 per request until maxItems is reached or the matches run out.
3. Each lot becomes one row; lots already seen in another search are skipped.
4. With includeDetails on, each lot's own page is fetched through the residential proxy and its VIN, colours, engine, registration, original market, description, gallery and view counts are merged into the row.
5. Rows are pushed to the dataset, ready to export as JSON, CSV or Excel.

⚙️ Input parameters

Field	Type	Default	Notes
startUrls	array	[]	Collecting Cars search, make/model or lot URLs
listingStage	select	live	live, comingsoon, active (live + coming soon), sold, all
lotTypes	array	["car"]	car, bike, watch, plate, part
saleFormats	array	all	auction, buyItNow, offer
makes	array	–	e.g. ["Porsche", "Ferrari"]; spelling and case are forgiving
models	array	–	e.g. ["911"]; must match the site's model name
keywords	string	–	Free text across titles, makes and models
countries	array	all	ISO-2 codes of where the lot is, e.g. ["GB", "DE", "AU"]
regions	array	all	The site's groupings, e.g. UK, EU, AU, US
driveSide	select	any	left / right
sellerType	select	any	private / trade / managed
noReserveOnly	boolean	false	Only lots offered without a reserve
priceMin / priceMax	integer	–	In the lot's own currency; bounds the sale price, or the current bid on a lot still running
yearMin / yearMax	integer	–	Model year
listedWithinHours	integer	–	Only lots published in the last N hours; 24 makes a daily new-listings feed
sortBy	select	relevance	newest, endingSoonest, recentlySold, priceHighest, priceLowest, bidHighest, mostBids
includeDetails	boolean	false	Fetch each lot's page for VIN, colours, description, gallery and more
maxItems	integer	1000	Total lots to return. The sold archive holds ~33,000, so set it deliberately
maxConcurrency	integer	10	Parallel lot-page requests when includeDetails is on
proxy	object	built-in residential	Used only for lot pages

For sale prices, set listingStage to sold. The default, live, matches what collectingcars.com/buy shows and keeps an exploratory run to a few hundred rows.

Example input

Every Porsche 911 sold in the UK or Germany, newest sales first, with full detail:

{
    "listingStage": "sold",
    "lotTypes": ["car"],
    "makes": ["Porsche"],
    "models": ["911"],
    "countries": ["GB", "DE"],
    "sortBy": "recentlySold",
    "includeDetails": true,
    "maxItems": 2000
}

📊 Output overview

One row per lot. The core fields come from the search index and are filled on every row. The detail fields (VIN, colours, engine, registration, original market, description, gallery, view and watcher counts) are null or empty unless includeDetails is on. Every field appears on every row, so CSV columns never shift between runs.

📦 Output sample

A real row, trimmed:

{
  "lotId": "91887",
  "lotUrl": "https://collectingcars.com/for-sale/1990-honda-nsx-1",
  "title": "1990 Honda NSX",
  "lotType": "car",
  "listingStage": "sold",
  "isSold": true,
  "saleFormat": "auction",
  "saleType": "buy_now",
  "currency": "GBP",
  "currentBid": 46750,
  "soldPrice": 50000,
  "bidCount": 58,
  "noReserve": false,
  "reserveMet": false,
  "auctionEndedAt": "2026-09-17T18:22:32.000Z",
  "soldAt": "2026-09-23T18:18:53.000Z",
  "publishedAt": "2026-09-10T10:00:08.000Z",
  "make": "Honda",
  "model": "NSX",
  "variant": "Base Model",
  "generation": "1st Gen - NA1",
  "year": 1990,
  "mileage": "67,265 Miles (Indicated)",
  "fuelType": "Petrol",
  "transmission": "automatic",
  "driveSide": "right",
  "country": "GB",
  "region": "UK",
  "location": "Surrey",
  "sellerType": "private",
  "priceGbp": 50000,
  "priceUsd": 66698,
  "priceEur": 58258,

  "vin": "JHMNA12600T000145",
  "engineSize": "3.0L V6",
  "bodyStyle": "coupe",
  "exteriorColour": "Red",
  "interiorColour": "Black",
  "originalMarket": null,
  "description": "KEY FACTS This Honda NSX is an iconic and highly engaging modern classic performance coupe. It is powered by a 3.0-litre V6 …",
  "imageCount": 132,
  "viewCount": 6632,
  "watcherCount": 318
}

This lot shows why saleType matters: bidding closed at £46,750 under the reserve, and the car sold afterwards for £50,000.

🗂 Key output fields

Group	Fields
State	listingStage (live / comingsoon / sold), isLive, isSold, saleFormat, saleType
Price	soldPrice, currentBid, bidCount, currency, buyNowPrice, priceGbp / priceUsd / priceEur (normalised by Collecting Cars, for comparing across currencies), priceBand
Reserve	noReserve, reserveMet, reserveLowered, soldPriceHidden
Timing	stageEndsAt (when a live auction closes), auctionEndedAt, soldAt, publishedAt
Vehicle	make, model, variant, generation, year, mileage, fuelType, transmission, driveSide
Location & seller	country, region, location, latitude, longitude, sellerType, sellerId, sellerBadges
Detail option	vin, engineNumber, engineSize, bodyStyle, exteriorColour, interiorColour, originalMarket, registration, description, images, imageCount, videoUrl, viewCount, watcherCount, sellerName, reservePrice, guidePriceLow, guidePriceHigh

❓ FAQ

Does it return the sale price on finished lots? Yes. 99.9% of the ~33,000 lots in the archive carry one. soldPriceHidden flags the roughly 20 where Collecting Cars withheld it.

Why is soldPrice sometimes higher than currentBid? Check saleType. On auction lots the car sold at the final bid, and the two match. On buy_now lots bidding closed under the reserve and the car sold afterwards at an agreed price, which is usually above the last bid. In our checks about 87% of sold cars were auction and about 12% buy_now.

Are the buyer's fees included? No. soldPrice is the price the lot sold for, before the buyer's fee that Collecting Cars adds on top.

Does the archive include cars that failed to sell? No. Every lot in it sold: the data has no "unsold" state, and 99.9% of archive lots carry a price. That makes it a record of results, not a source for sell-through rates.

How do I get the whole archive? Set listingStage to sold, tick every lot type, and set maxItems to 35000. A test run returned all 33,289 sold lots in 62 seconds, from 134 search requests. With includeDetails on, add one lot-page request per lot.

Why is originalMarket empty on so many cars? It is only set on imports. A car sold new in the country it is being sold from has no separate original market to record.

Does it log in or use a browser? Neither. The core fields come from the site's public search data, and the detail option reads the public lot page.

How much will a run cost? $3 per 1,000 lots, whether or not the detail option is on. The whole sold archive comes to about $100.

💬 Support

• Issues tab on the Actor page
• Email: muhamed.didovic@gmail.com

🛠 Additional services

• Custom output shape, extra fields, or a one-off dataset: muhamed.didovic@gmail.com
• Need the same for another collector-car auction house (Bring a Trailer, Cars & Bids, Historics, Catawiki)? Drop an email.
• For API access (no Apify fee, just a usage fee for the API): muhamed.didovic@gmail.com

🔎 Explore more scrapers

See other scrapers at memo23's Apify profile, including the PistonHeads classifieds-and-auctions scraper and the Car & Classic scraper, plus job boards, real estate and more.

---

🤖 For AI Agents & LLM Apps

• Deterministic input. Every filter is plain JSON; makes are matched against the site's own list and a name that matches nothing stops the run instead of returning zero rows.
• Stable schema. Every field is present on every row (null when absent), so downstream parsing needs no defensive key checks.
• Bounded cost. maxItems caps the number of rows exactly, and a pasted URL the Actor cannot read is refused rather than widened into a full-catalogue run.
• Unambiguous prices. saleType says whether soldPrice is the final bid or a post-auction sale, and currency plus the normalised priceGbp/priceUsd/priceEur make lots comparable across markets.
• Grounding-friendly. lotUrl gives every row a citable public source; soldAt, publishedAt and scrapedAt support freshness reasoning.

---

⚠️ Disclaimer

This Actor is an independent tool and is not affiliated with, endorsed by, or sponsored by Collecting Cars Limited, Watch Collecting, or any of their subsidiaries or affiliates. All trademarks mentioned are the property of their respective owners.

The scraper accesses only publicly available listing data on collectingcars.com and watchcollecting.com: no authenticated endpoints, member-only features, or content behind a login. It does not place bids. Users are responsible for ensuring their use complies with Collecting Cars' Terms of Service, applicable data-protection law (GDPR, CCPA, etc.), and any contractual obligations of their own organisation.

---

SEO Keywords

collecting cars scraper, collectingcars scraper, scrape collectingcars.com, collecting cars api, Apify collecting cars, collecting cars sold prices, car auction results scraper, classic car auction data, collector car auction prices, car auction sold prices dataset, classic car valuation data, supercar auction results, live car auction bids, car auction archive, classic car market data, porsche 911 auction results, watch collecting scraper, watch auction results scraper, number plate auction data, bring a trailer alternative, car and classic alternative, pistonheads auctions alternative