How much does VIN data cost per vehicle?

The cost of VIN data per vehicle follows the data's own basis, not a flat rate. Faces that come from VIN's own database — the decode, the factory specs, the live-listings corpus, and VIN's own values model — carry no per-pull cost. Faces metered against a cheap public source, like the recalls check, are priced thin. And real-money faces licensed per lookup — book values, the OEM build record, the window sticker, and the full history report — are priced through at what VIN pays upstream. Each face posts its own license and its own price, and one VIN's whole record settles as a single typed bundle. There is no per-VIN rate card here to quote, because the honest answer is a structure, not a number.

That structure is the point. The failure mode this replaces is procurement theater: coverage you discover after the contract, a price that depends on who is asking, and "contact sales" where a buyer needed a price sheet. The catalog reads the same to every buyer. What follows is the cost model, tier by tier.

The three cost tiers

Tier one — zero-marginal own-database faces. The decode, the factory specifications, the live-listings spine, and VIN's own values model are computed from data VIN already holds. There is no third party to pay per lookup, so there is no per-pull cost to pass through. At business scale these faces are the base of the catalog: read them, quote them, license them at volume — the marginal cost of one more VIN's decode is effectively nothing, and the pricing reflects that rather than inventing a fee.

Tier two — metered-cheap faces. Some faces read a public source at low, real cost. The open-recall check reads NHTSA campaign data; the taxes calculation runs against public rate tables. These are priced thin — a small metered cost per lookup, because there is a real (if minor) cost to the pull. They are not free like tier one, but they are nowhere near the licensed tier.

Tier three — real-money licensed faces. Book values (the KBB / Black Book / J.D. Power overlays), the OEM build data, the original window sticker, and the full vehicle history report are each licensed from an upstream provider, per pull, at real money. VIN does not absorb that cost or give it away — it is priced through: the buyer pays what the data actually costs, per face, with the license terms stated alongside. This is the deepening tier, and it is where per-VIN cost concentrates.

Why the tiers, not a flat price

A flat per-VIN price would either overcharge for the free faces or undercharge for the licensed ones — and it would hide which is which. The tiered model exists so the cost is legible: a buyer pulling only decode and specs at volume pays for infrastructure, not per-VIN licensing; a buyer pulling full history and book values on every VIN is paying real upstream cost and can see exactly that. The price tracks the data's origin, so the invoice is explainable line by line.

This also governs the two flagship faces buyers ask about most. The full history report and the licensed valuation overlays are the two most expensive things a VIN's record can contain — both tier three, both priced through. What a car is worth has a free base number and a paid, build-level accurate reading; that ladder reads at values.vin. What a car's history holds is the paid report itself; that reads at vhr.vin. The catalog composes their prices as each door states them — it never undercuts a member's own gate, and a bundle that contains the history report credits it by that member's own rule, restated here, never re-priced.

How it settles — the typed per-VIN bundle

One VIN's whole record prices as a single typed bundle. At data.vin/{VIN}, the coverage-and-quote manifest states which datasets hold rows for that VIN and composes the full record's data into one machine-readable offer: a posted amount, an x402 settlement envelope, and an expiry. The manifest is free on every VIN; the bundle is what settles when a buyer takes it. Because the offer is typed and posted, the worst-case cost is on paper before the first byte moves — and an autonomous purchasing agent can read and settle it without a sales call, which a "contact us" quote makes impossible.

At volume, the catalog licenses on paper: metered plans (Starter, Growth, Scale) and bulk / firehose priced through negotiation, with the free tier — read the catalog, quote any VIN — always at zero. One caveat is stated honestly: the resale bundle settles only once the upstream data license names catalog resale and the bundle envelope; until that gate clears, the manifest still quotes today, free. Keys, quotas, SDK, and MCP integration are a separate concern handled at apis.vin — this page prices what you integrate, it does not sell the pipe.

So: how much does VIN data cost per vehicle? As much as its data actually costs to produce, and no more — zero for the own-database faces, thin for the metered ones, priced-through for the licensed ones, each priced per face and settleable as one bundle. To price a specific VIN's whole record, read the manifest at data.vin.