Hepsiburada API for product, review and search data

Hepsiburada offers a merchant integration API, and it does what a merchant needs: managing a seller's own listings, stock, prices and orders. It covers that seller's store data only. It is not a way to read another seller's listing, a competitor's reviews, or search results across the marketplace.

The Roketfy Hepsiburada API covers the public side instead. It reads the same product, review, question, search and category pages a shopper sees and returns them as JSON, for any product or seller on the site. You do not need a Hepsiburada seller account, and you do not need to be based in Turkey.

`/v1/hepsiburada/product-detail` returns the full product record from a product URL or SKU: specs, variants, images, ratings, and every seller on the listing with their prices and buybox position. One call shows both the product itself and the competition on it.

Hepsiburada has two product URL forms, and the difference affects cost. A URL ending in -p-<sku> names a specific variant and costs 1 credit. A URL ending in -pm-<productId> names a product rather than a variant, so it has to be resolved to a SKU through the product page first, and the call costs 2 credits. If you store SKUs or -p- URLs in your own system, every call stays at 1 credit.

An optional `merchant` parameter scopes the listing to one seller by name. When the URL you pass already carries a `magaza` parameter, the seller is taken from it.

`/v1/hepsiburada/reviews` returns customer reviews page by page: the review and its photos, per-aspect stars, the seller it was bought from, plus star, seller and keyword facets and Hepsiburada's own AI summary of the reviews. `/v1/hepsiburada/questions` returns the published buyer questions on a listing with each seller's answer, when it was asked and answered, and the like and dislike counts.

Both endpoints cost 1 credit per page, or 2 for a -pm-<productId> URL, and both support batch tasks. Instead of looping over pages yourself, you submit one task and the Data Engine collects the pages for you at the same per-page cost.

Review and question text is in Turkish, as written by buyers. The JSON structure around it is the same for every product, so it can go straight into a translation or sentiment step.

Yes. `/v1/hepsiburada/search` returns the results for a search term in storefront order: every card's page position, price, campaign price, seller, badges and rating, plus the facets the listing offers and whether the term was typo-corrected. `/v1/hepsiburada/category-products` returns the same cards for a category page instead of a search term. Both support batch tasks.

`/v1/hepsiburada/search-suggestions` returns Hepsiburada's own autocomplete for a term: the completions it offers, plus the categories, brands and sellers it surfaces alongside them. It is a quick way to find the Turkish search terms shoppers actually use before you run full searches.

A self-built Hepsiburada scraper means fetching pages, getting past blocking, parsing the markup and fixing the parser each time the site changes. For a team outside Turkey there is the added work of understanding a Turkish-language storefront well enough to know which fields matter.

A structured API removes that work. Each endpoint returns JSON with documented parameters for one page type, so integration is a single HTTP request with a Bearer token. You trade some flexibility for not owning a scraper, which is usually the right trade when Hepsiburada is one data source among several.

Every Hepsiburada endpoint costs 1 credit per successful call, and each page of reviews, questions, search or category results counts as one call. The one exception is the -pm-<productId> URL form on product detail, reviews and questions, which costs 2 credits because of the extra resolution step.

Credits are bought as pay-as-you-go packages, and failed requests are not charged. All six endpoints are also available as MCP tools, so an AI agent can look up a Hepsiburada product, its reviews or search results with the same token and credit balance. Full parameters are in the docs at data.roketfy.com/docs.