Korean car marketplace templates

A codebase, not a subscription. You pay once, you get a ZIP, you put your name on it and you host it wherever you already host things. There is no licence server to phone home to, no per-listing fee, and no way for anyone here to switch your site off.

What makes it worth buying rather than building is the half nobody sees: the Korean fields are already translated on the way in, the price unit trap is already handled, sold cars already remove themselves, and the request budget is already managed by a cache layer. Those four things are where a from-scratch build loses its first month.

The two are not a good version and a better version. They are for two different situations, and picking by price is how people end up with the wrong one.

The honest split. If more than two rows on the right-hand side describe you, buy the Next.js starter even though it costs twice as much, because converting the WordPress build into it later costs more than the difference.

Both templates assume you have a free API key already, which takes a minute and costs nothing. From there the WordPress route is an upload and a settings screen; the Next.js route is an install, one environment variable and a deploy.

Nothing in either one needs a database of cars. That is the part people expect and are surprised by: your site stores no inventory, because the inventory is fetched and cached. There is nothing to seed, nothing to import nightly, and nothing to go stale.

This is the thing that decides whether a self-hosted marketplace is cheap or expensive to run, and it is already solved in both templates. A naive build calls the API once per visitor per page view, which turns a hundred visitors into several thousand requests and pushes a free key into a paid plan on its first good day.

Both templates cache instead. WordPress uses transients; the Next.js starter caches rendered vehicle pages and revalidates them on a timer. The catalogue stays current to within minutes, and the request count tracks how many distinct cars and searches exist rather than how many people looked at them.

The licence is single-project: one template, one site, yours forever, modify it as much as you like. A second site is a second licence. You may not resell the template itself as a template, which is the only restriction that matters in practice.

Upstream changes are handled on our side of the API rather than yours. When Encar.com alters a payload — and it does — the API absorbs it and the shape your template reads stays the same. That is the actual argument for this over a scraper: a scraped build breaks on someone else's schedule and you find out from a customer.

The phrase people search for is “AutoTrader clone”, and the market that answers it sells a PHP bundle with a screenshot gallery and no data behind it. The site arrives empty and the hard part — getting real, current, correctly translated inventory into it — is left as an exercise, usually with a suggestion that you scrape it.

These templates are the opposite trade. The front end is deliberately ordinary; the value is that it is already attached to a maintained feed of 200,000+ live Korean listings with the Korean fields translated and the sold ones removing themselves. You are buying the integration, and the marketplace comes with it.