Data Engine API reference: every scraping endpoint

Full Trendyol product record from a product URL or SKU: attributes, variants, ratings, the buybox winner and every other seller with their prices and stock.

Customer reviews for a Trendyol product, page by page: rating breakdown, aspect tags with sentiment, reviewer body measurements and photos, plus Trendyol's own AI summary.

Answered buyer questions on a Trendyol listing, page by page: the question, the seller's answer, how fast they replied, and the subject tags the product's questions fall under.

Trendyol's own category leaderboards — best sellers, most viewed, most favourited or most rated — with each product's rank, price, rating and seller.

Trendyol search results for a term, in storefront order: every product's page position, price, rating, seller and badges — and which rows are paid placements rather than organic results.

Everything listed in a Trendyol category, in storefront order — the same rows a search returns, for a category page instead of a search term.

Everything one seller has listed on Trendyol, in storefront order — the same rows a search returns, scoped to their store instead of a term or category.

Trendyol's own autocomplete for a search term: the completions it offers, plus the categories, campaigns and sellers it surfaces alongside them.

What Turkey is searching for on Trendyol right now — the ranked term list behind the storefront's empty search box, with the terms Trendyol is highlighting.

Full Hepsiburada product record from a product URL or SKU: specs, variants, images, ratings, and every seller on the listing with their prices and buybox position. Costs 1 credit — or 2 for a …-pm-<productId> URL, which names a product rather than a variant and has to be resolved to a SKU through the product page first.

Customer reviews for a Hepsiburada product, page by page: the review and its photos, per-aspect stars, the seller it was bought from, plus the star/seller/keyword facets and Hepsiburada's own AI summary. Costs 1 credit per page — or 2 for a …-pm-<productId> URL, which has to be resolved to a SKU through the product page first.

Published buyer questions on a Hepsiburada listing, page by page: the question, each seller's answer, when it was asked and answered, and the like/dislike counts. Costs 1 credit per page — or 2 for a …-pm-<productId> URL, which has to be resolved to a SKU through the product page first.

Hepsiburada search results for a 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.

Everything listed in a Hepsiburada category, in storefront order — the same cards a search returns, for a category page instead of a search term.

Hepsiburada's own autocomplete for a term: the completions it offers, plus the categories, brands and sellers it surfaces alongside them.

Full Amazon product page for an ASIN: title, pricing, images, specifications and description, from the marketplace you name.

Buyer reviews for an Amazon listing, page by page: rating, title, text, date, author and helpful votes, plus the review count and seller the page header states. Requires your own signed-in Amazon cookies, passed as `cookies` — the pager answers a session and nothing else, so a call without one is refused rather than answered with an empty page.

Every seller offering an ASIN, with price, shipping, condition, fulfilment and which one holds the Buy Box.

Amazon search results with listings and their ranking. Category pages use the same structure, so a category term works here too.

The complete HTML of any Amazon URL, rendered from the location you name. Returns `{ html }` rather than structured JSON; capped at 4 MB upstream.

An n11 listing in full: price, stock, images, attributes, category path, its variants and their stock, the seller with its feedback scorecard, and the product's rating breakdown — including the numeric product id the review endpoint takes.

Buyer reviews for an n11 product: the page of reviews with their photos, votes, seller and masked buyer, plus the product's rating breakdown, n11's own AI summary and the tag facets that product offers.

n11's own autocomplete for a term: the completions it offers, plus the categories and seller stores it surfaces alongside them.

A product as Akakçe's comparison page describes it: name, brand, images, spec table, the price range across every seller, the 30-day low, follower count and a handful of offers with real outbound seller URLs. For the full, sortable, filterable seller list use akakce/offers.

Every seller Akakçe lists for a product — price, shipping cost, stock, delivery estimate and instalment offers — sorted by price, price with shipping, or seller rating, and narrowable to specific marketplaces.

The same seller list as akakce/offers, with each offer's own link on the seller's site or marketplace resolved — the page the listing actually lives on, not an Akakçe page. Takes the same sort and filters. Charged only when at least one link comes back.

How this product's cheapest price has moved: one row per day it actually changed — the date, the new price, and how much and which way it moved — from Akakçe's own price-change log, newest first.

Akakçe search results for a term: every matching product's own lowest price, seller count, follower count and image. Narrow with filterIds from akakce/search-filters, or a price range.

The facet menu behind an akakce/search term — brand, feature, dimension, color and seller filters, each with the id akakce/search's own filterIds takes. Facets are per search term, never a fixed list: read this before filtering.

Akakçe's own search-box dropdown for a partial term: products it matches directly (each with the prCode every other Akakçe endpoint takes) plus broader suggested search terms.

Everything listed in an Akakçe category, in the same shape akakce/search returns — every product's own lowest price, seller count and follower count. Identify the category by its numeric ctCode, or its URL (costs one extra credit to resolve).

Google search results for a query, localised by domain, language and country.

The same query run through Google's AI mode rather than the classic result page.

Google Maps places for a query: a ranked list with ratings, addresses, place ids and CIDs.

One Google Business profile in full: categories, hours, attributes, rating distribution and review topics. Identify it by name, or by `cid:<id>` / `place_id:<id>` as the keyword.

Google reviews for a place — text, rating, timestamps, images and owner responses. Up to 4490 in one call.

Google Shopping results for a query — up to 40 products a page, with prices.

One Google Shopping product: every seller offer, variants, specs, reviews and videos. Identify it by catalog_id + q (the durable pair), by page_token, or by product_id for a leaner response.

The continuation endpoint for a shopping product's remaining sellers. This is where a next_page_token belongs — the product endpoint refuses it.

Google Flights itineraries and prices for a route and date.

Google Hotels availability and nightly prices for a place and a pair of dates.

One Google Hotels property in full, for the dates you name.

Google News, driven by exactly one of: a query, a topic, a section, a story cluster, a publisher, or a Knowledge Graph entity.

Google Trends interest over time for a term in a region, plus related queries.

What is trending on Google right now. Takes no parameters of its own.

One prompt to Google Gemini, answered as markdown with its sources. Stateless — no conversation is kept. Costs 25 credits, not 10.

Google Play listings by search term, by category, or by chart (top free, top paid, top grossing).

One Google Play app: description, ratings, screenshots and metadata.

Paginated user reviews for a Google Play app.

App Store search results for a keyword, ranked as the store returns them.

Store App One App Store listing: title, developer, category, rating, price and metadata.

User reviews for an App Store app — title, text, rating and version.

Find businesses on Trustpilot by company name or category. Returns the domain each review page is addressed by.

Trustpilot reviews for a business domain — rating, title, text, reviewer and any company reply.

Find hotels, restaurants and attractions on Tripadvisor by name or category. Returns the url_path the reviews endpoint takes.

Tripadvisor reviews for a business — rating, title, text, visit type and owner responses. Up to 4490 in one task.

Walmart search results for a query, priced against the store or ZIP you name.

A Walmart product by item id, priced against the store or ZIP you name.

A store-scoped page fetch: which Walmart store a ZIP resolves to, and that store's own details.

YouTube search results: videos, channels, playlists and shorts, plus the related searches and 'people also watched' shelves.

GPT Chat One prompt to ChatGPT, answered as markdown and HTML with its sources, plus whatever shopping results, ads and web searches it used. Stateless — no conversation is kept. Costs 25 credits.

Fills missing product attribute values by researching the web — the brand's own site, Trendyol and Hepsiburada, a broader search, then Google's AI-mode summary — and reconciling what it finds with an LLM. Multi-step per product; reachable only through POST /v1/tasks. Flat 25 credits per product regardless of how many stages ran.