FFetchSmith

Guides / 2026-09-09

Every Shopify store's catalog is public JSON — no login, no browser, no API key

"/products.json and /meta.json expose a Shopify store's full catalog, prices, variants and currency to anyone, paginated, with zero authentication. Here's the shape of it and what most scrapers still get wrong."

Shopify storefronts ship their entire product catalog as plain JSON, at a URL every store has, with no authentication and no rate-limit key. If you can load the store's homepage in a browser, you can fetch its full catalog with curl.

The endpoint

Every Shopify store exposes:

GET https://<store-domain>/products.json?limit=250&page=1

This works identically on *.myshopify.com backends and on custom storefront domains (www.allbirds.com, gymshark.com, whatever the merchant put on their DNS) — Shopify doesn't gate the endpoint behind the domain type. limit caps out at 250 per page; paginate with page=2, page=3, ... until a page comes back with an empty products array, which is your signal to stop (Shopify doesn't return a total count or a Link header for this endpoint, so "empty page" is the only reliable end-of-catalog marker).

Scope to a single collection instead of the whole store by inserting the collection handle:

GET https://<store-domain>/collections/<handle>/products.json?limit=250&page=1

Each product in the response carries variants[] (SKU, price, weight in grams, availability, option values like size/color) and images[]. There's no currency field on the product payload itself — for that you need a second, separate request:

GET https://<store-domain>/meta.json

which returns the store's currency code among other shop metadata. One extra request per store, cached once, gets you currency for every product from that store without re-fetching it per item.

What this data doesn't tell you, and why that's fine

/products.json has no explicit "on sale" flag. What it does have is price and compare_at_price on each variant — when compare_at_price is set and higher than price, that variant is discounted. Roll that up across a product's variants (min price vs. min compare-at price) and you get a reliable isOnSale boolean without ever touching a rendered page.

Two failure modes worth handling explicitly instead of letting them look like empty catalogs: - Not a Shopify store, or the endpoint is disabled. Some merchants disable /products.json (Shopify allows opting out via a theme/app setting). The request itself fails or 404s — that's a fetch error, not "the store has zero products," and the two should never be reported the same way. - Genuinely empty store or collection. The request succeeds and returns {"products": []}. That's a real empty result, distinct from the case above.

We hit both of these building Shopify Products Scraper and initially conflated them into one silent "0 results" — fixed by classifying "fetch failed" vs. "fetched fine, zero products" vs. "your onlyAvailable filter removed everything" into separate status messages, so a run's outcome is legible without opening the dataset.

Practical takeaways

  • No login, no API key, no headless browser — /products.json and /meta.json are two plain HTTP GETs.
  • Paginate on limit=250&page=N until you get an empty products array; there's no total-count field to pre-compute the number of pages.
  • Fetch /meta.json once per store for currency, not once per product.
  • Derive "on sale" from price vs compare_at_price across variants — Shopify doesn't hand you a boolean, but it hands you everything needed to compute one.
  • Distinguish "this isn't a working Shopify catalog" from "this is a working, empty catalog" — they look identical in a naive implementation and mean opposite things to whoever's reading the run.

This is exactly what Shopify Products Scraper does under the hood: no per-run start fee (some competitors charge ~$0.10 just to start before any data is delivered), charged only per product actually returned.

Built by FetchSmith — HTTP-only Apify Actors, AI-assisted development, disclosed.