OpenLane auction data: coverage, lineage, and programmatic access

OpenLane operates as a digital remarketing platform in Europe, bringing together vehicle supply that had previously been distributed across several separately branded businesses. For anyone consuming the data, the useful thing to understand is not the corporate history but what kind of vehicles flow through it — because remarketing supply behaves very differently from consumer trade-in supply.

Remarketing supply versus trade-in supply

Wholesale platforms broadly split by where their stock comes from, and it shows up clearly in the data.

The cohort property is the one worth building around. When a leasing company returns forty identically specified Passats at the end of a contract, you get a set of near-duplicate vehicles that differ mainly in mileage and condition. For a pricing model that is unusually clean signal — it isolates the mileage and condition variables from everything else. For a marketplace frontend it is a UX problem, because forty near-identical results is a bad search page.

A note on ADESA and CarsOnTheWeb

Two legacy names come up in searches: ADESA Europe and CarsOnTheWeb. Both are associated with the European remarketing business that now trades as OpenLane. We do not maintain historical lineage claims we cannot verify, and we would suggest treating any confident account of which brand became which — including in tools that surface old documentation — as something to check against OpenLane's own material rather than take at face value.

The practical implication for a developer is narrow but real: if you have historical data keyed on an old brand identifier, do not assume it maps cleanly onto current OpenLane lot ids. Verify before you join on it.

What the listings carry

Remarketing listings are generally strong on documented condition, because the vehicles were inspected under a contractual standard rather than described by whoever is selling them. Expect the specification and history fields to be well populated, and the damage description to be structured around defined wear categories rather than free text.

That said, fill rates still vary by field and by consignor, and this is worth measuring rather than assuming. A field that is 95% populated for one consignor's stock and 40% populated for another's will average out to something misleading.

Pulling OpenLane inventory

OpenLane sits under the site slug openlane, addressed identically to every other source:

# Recent low-mileage OpenLane stock in the Netherlands
curl -s "https://api.thecarapi.com/api/search?site=openlane&country=NL\
&year_from=2021&kilometers_to=80000&sort=price_low&limit=24" \
  -H "X-API-Key: $API_KEY"

# Detail for one lot
curl -s "https://api.thecarapi.com/api/auction/openlane/11125938" \
  -H "X-API-Key: $API_KEY"

# How the price on that lot has moved since we first saw it
curl -s "https://api.thecarapi.com/api/auction/openlane/11125938/price-history" \
  -H "X-API-Key: $API_KEY"

Because remarketing stock arrives in cohorts, price history on OpenLane lots is more informative than on one-off consumer stock: you can watch how a set of near-identical vehicles clears, which is close to a controlled experiment in what the market will pay. The price history guide covers how to read that series properly.

Where OpenLane fits against the other sources

Nobody should be sourcing from one platform. The reason to hold several is that they cover genuinely different supply:

• OpenLane — fleet and lease returns. Predictable, documented, cohorted. Best signal for pricing models.
• Auto1 — broad consumer and mixed supply, largest volume, widest variety. See the Auto1 access options.
• Schadeautos and Copart Germany — damaged and salvage, a different buyer and a different funnel entirely.
• eCarsTrade — ex-lease and ex-rental fleet stock, overlapping with OpenLane but with its own consignor base.
• Encar — Korean domestic supply, a separate market with separate price dynamics.

A query filtered to one source answers "what is available here". A query across all of them answers "what is this car worth right now", which is the question with commercial value. That comparison only works if the sources have been normalized onto one schema first — see vehicle data normalization for why that is harder than it sounds.