The database
nobody asked
for. And the one
everyone needs.

There's a moment, familiar to anyone who has spent time hunting for a specific car, when the internet fails you completely. You've found the listing. The photos look right. The year checks out. But the specs — displacement, original output, gearbox type, whether the car left the factory with the correct colour — are either wrong, missing, or lifted from some forum post written in 2007 by someone who was guessing. You close the tab. You're no further along than when you started.

That's the problem we set out to fix. Not with enthusiasm, exactly — more with a kind of low-grade irritation that eventually crosses into purpose. The internet has an extraordinary amount of information about cars. What it doesn't have, or didn't until recently, is organised, verified, cross-referenced information — the kind you can actually use when you're trying to make a decision, settle an argument, or understand something properly.

The automotive world is unusual in how much it cares about its own history while simultaneously doing so little to preserve it accurately. Manufacturers issue press releases. Magazines run road tests. Enthusiast forums accumulate decades of posts, half-remembered dyno figures, and disputed production numbers. But nobody, for a long time, sat down and said: let's make a single place where all of this is checked, dated, and findable.

Murray was talking about something specific — the mythology that accumulates around iconic cars, the stats that get rounded, the figures that get confused with later variants — but he could have been describing the state of automotive information online at large. A 1973 Carrera RS doesn't have 230 horsepower. It never did. But that number has appeared in enough places that it now competes with the correct one. When you're trying to buy a car, or write about one, or simply understand one, that kind of noise is more than annoying. It's a genuine obstacle.

The archive is, at its most basic, a database. Each entry contains the make, model, year, variant, powertrain specifications, dimensions, production numbers, and — where we have them — factory documents or period test results that verify the figures. There are currently over 3,800 entries, spanning from the 1886 Benz Patent-Motorwagen to cars announced last week. Every entry has been touched by a human. None of them are generated.

But a database is only as useful as its interface, and what we've tried to build is something that doesn't feel like querying a spreadsheet. The comparison tool lets you stack any three cars against each other across every spec simultaneously. The era browser lets you filter by decade. The production filter — genuinely useful if you're trying to understand scarcity and value — lets you sort by how many were actually built. These aren't features we added because they seemed like good features. They're things we needed ourselves and couldn't find anywhere else.

People come to the archive for different reasons. Some are researchers. Some are journalists. A lot of them are people trying to buy a car — and that, it turns out, is where having reliable data matters most. The secondhand car market runs on information asymmetry. The seller knows more than the buyer, or the buyer knows more than the seller, and either way, the party with worse data is at a disadvantage. An archive that tells you exactly what a car should weigh, what it should make, what was standard and what was optional — that's not just interesting. It's protective.

We hear versions of this constantly. Someone finds a listing for a car they want, pulls it up in the archive, and immediately spots that the stated output is 20 horsepower higher than the factory figure — a classic sign that someone has claimed a tune that may or may not exist. Someone else is looking at a limited edition variant and wants to know whether the production number the seller is quoting is plausible. These aren't edge cases. They're the everyday reality of buying a car with any kind of history.

We don't tell people what to buy. That's not what this is. But we give people the information they need to ask the right questions — and in a market where the wrong answer can cost tens of thousands, that's not nothing.

There's something about maintaining an archive that resists a clean description of purpose. It's not exactly preservation — the cars exist whether we document them or not. It's not exactly journalism — we're not breaking stories. It's closer to what a good reference library does: it makes existing knowledge findable, reliable, and free. That feels old-fashioned. It also feels, right now, necessary.

We add new entries every week. We revise old ones when new documents surface — factory build sheets that contradict the received wisdom, period road tests that clarify a disputed figure, updated homologation records. History turns out not to be finished. The number of production units on a rare 1960s sports car can still be revised when a previously unknown chassis turns up in a barn in southern France. That kind of update matters, and we make it.

There's no algorithm deciding what gets added or how. There's no engagement metric being optimised. The archive grows because there are cars worth documenting and people willing to do the work of documenting them correctly. That's it. That's the whole model.

If that sounds like something you could use — whether you're buying, selling, writing, researching, or just trying to settle an argument — the archive is open. No account required. Start wherever you like.