Trendyol API & Scraper
The Trendyol API returns product data from Turkey's largest e-commerce marketplace as clean JSON.
🤖 Using an AI assistant? Copy this link into ChatGPT / Claude / Cursor — it reads every endpoint and parameter instantly and tells you if this API fits your use case.
The primary product/detail endpoint returns a product's content id, title, brand, price (current, list, discount, coupon, TY-plus) and campaign, and you can pull product/reviews, similar products, run a search with filters, browse a category, pull a seller's products and info, and get suggestions. It is built for price intelligence, catalog enrichment and Turkish e-commerce analytics that need Trendyol data without a scraper. One ReefAPI key, one shared credit pool, the standard envelope.
Trendyol ids, price formats and the numbers behind them
Two things trip up a first integration here. Every price is returned twice, once as a locale-formatted Turkish string and once as a number, and Trendyol uses four different numeric ids that all look alike. The values below were read from live calls on the iPhone 15 128 GB listing (content_id 762254881) and a laptop search on 2026-08-27.
| Field | Measured value | What it means |
|---|---|---|
| price.current / current_value | "49.999 TL" / 49999 | Turkish formatting: the dot is a thousands separator. The string reads as 49.99 in most parsers. Always use current_value. |
| price (search row) | "74,90 TL" / price_value 74.9 | In the same API a comma is the decimal separator. The two formats coexist, which is why the *_value fields exist. |
| content_id | 762254881 | The digits after -p- in a product URL. This is what product/detail, product/reviews and product/similar take. |
| product_group_id | 580170099 | The parent group a product's colour and storage variants belong to. Not interchangeable with content_id. |
| seller_id / merchant_id | 968 (Trendyol), 275331 (MediaMarkt) | The digits after -m- in a store URL. Feeds seller/info and seller/products. Search rows expose it as merchant_id. |
| category.id | 5473 ("IOS Cep Telefonları") | The digits after -c in a category URL. Feeds category/products. Detail also returns the full hierarchy string. |
| brand filter id | 102323 = LENOVO, 101606 = ASUS, 101849 = HP | The brand parameter takes a numeric id, never a brand name. Get ids from search/filters, facet "Marka". |
| currency | "TRY" | Every price on every action is Turkish lira. There is no currency parameter. |
Paging is deliberately inconsistent between actions, because Trendyol's own endpoints are. search, category/products and seller/products are 1-based at roughly 24 products per page with max_pages capped at 10, while product/reviews is 0-based with page_size up to 50. summary.total_pages on reviews is computed for the page_size you sent: the same product reported 313 pages at page_size 3.
Real request and response JSON
Captured from the indexed primary action, search, on .
{
"method": "POST",
"url": "https://api.reefapi.com/trendyol/v1/search",
"headers": {
"x-api-key": "$REEF_KEY",
"content-type": "application/json"
},
"body": {
"query": "iphone"
}
}{
"ok": true,
"meta": {
"api": "trendyol",
"endpoint": "search",
"mode": "live",
"latency_ms": 1121.7,
"record_count": 36,
"bytes": 188571,
"cache_hit": false,
"version": "0.1",
"total_count": 100000,
"page": 1,
"pagination": {
"page": 1,
"has_more": true
},
"charged_credits": 2
},
"data": {
"results": [
{
"content_id": 985256821,
"product_group_id": 750312515,
"title": "iPhone 17 Pro Max 256GB Gümüş",
"brand": "Apple",
"url": "https://www.trendyol.com/apple/iphone-17-pro-max-256gb-gumus-p-985256821?boutiqueId=689770&merchantId=968",
"category": "IOS Cep Telefonları",
"merchant_id": 968,
"image": "https://cdn.dsmcdn.com/mnresize/400/-/ty1788/prod/QC_PREP/20251118/16/630c73c3-2c09-3158-9ad1-80cff418884c/1_org_zoom.jpg",
"price": "123.499 TL",
"price_value": 123499,
"list_price": null,
"list_price_value": null,
"rating": 4.611111111111111,
"comment_count": 180,
"free_cargo": false,
"fast_delivery": null,
"flash_sale": null
},
{
"content_id": 985256825,
"product_group_id": 750312515,
"title": "iPhone 17 Pro Max 256GB Kozmik Turuncu",
"brand": "Apple",
"url": "https://www.trendyol.com/apple/iphone-17-pro-max-256gb-kozmik-turuncu-p-985256825?boutiqueId=689770&merchantId=968",
"category": "IOS Cep Telefonları",
"merchant_id": 968,
"image": "https://cdn.dsmcdn.com/mnresize/400/-/ty1761/prod/QC_PREP/20250922/16/5d65ec4d-08cd-32aa-979f-f33fb44a989d/1_org_zoom.jpg",
"price": "119.999 TL",
"price_value": 119999,
"list_price": null,
"list_price_value": null,
"rating": 4.597938144329897,
"comment_count": 194,
"free_cargo": false,
"fast_delivery": null,
"flash_sale": null
},
{
"content_id": 762254881,
"product_group_id": 580170099,
"title": "iPhone 15 128 GB Mavi",
"brand": "Apple",
"url": "https://www.trendyol.com/apple/iphone-15-128-gb-mavi-p-762254881?boutiqueId=689770&merchantId=968",
"category": "IOS Cep Telefonları",
"merchant_id": 968,
"image": "https://cdn.dsmcdn.com/mnresize/400/-/ty1609/prod/QC/20241202/09/d1662cea-116d-3e2c-9a7a-66c6204ea0a3/1_org_zoom.jpg",
"price": "57.249 TL",
"price_value": 57249,
"list_price": null,
"list_price_value": null,
"rating": 4.643058161350845,
"comment_count": 2132,
"free_cargo": false,
"fast_delivery": null,
"flash_sale": null
}
],
"query": "iphone"
}
}What the Trendyol API does
| Action | Description | Concrete use case | Key params |
|---|---|---|---|
| product/offers | Selected seller and published competing seller offers for a product. Selected seller first, then competitors ordered by price. Stock is null where unpublished. | Pricing teams call product/offers to get selected seller and published competing seller offers for a product. | content_id, url, limit |
| product/variants | Published catalog variants and barcodes. Catalog prices can differ from the selected merchant. selected_offer_price is populated only for the matching selected item_number; other merchant-specific variant prices remain null. | Marketplace operators call product/variants to get published catalog variants and barcodes. | content_id, url |
| product/description | Long product description as HTML, plain text and image URLs. Image-only descriptions are supported. An existing product without a description returns has_description=false. Treat returned HTML as untrusted data before displaying it. | Catalog enrichment teams call product/description to get long product description as HTML, plain text and image URLs. | content_id, url |
| product/questions | Published answered product questions, with seller replies and zero-based pagination. total_count describes the returned question collection; content_total_count is a separate broader source summary. | Retail analysts call product/questions to get published answered product questions, with seller replies and zero-based pagination. | content_id, url, page, page_size |
| product/coupons | Publicly visible coupons for the selected product and seller. Includes discount type, minimum spend, expiry and conditions. Anonymous visibility does not confirm personal eligibility or redemption; coupons are not subtracted from the product price. | Pricing teams call product/coupons to get publicly visible coupons for the selected product and seller. | content_id, url |
| product/detail | full product by `url` OR `content_id`: title/brand/price(current+list+discount%/coupon/ty+)/campaign(+campaigns[])/rating+comment_count/in_stock/seller(+score)/other_offers[] (EVERY competing seller on the same listing WITH its store id, store name, seller score, price and campaign — seller-level price comparison in this one call)/category/images/attributes(specs)/variants. The dual-use influencer module. | Marketplace operators call product/detail to get full product by `url` OR `content_id`. | url, content_id |
| product/reviews | Paginated customer reviews by `content_id` OR `url`, in the order you ask for. `sort=newest` returns the most recently posted reviews first and keeps that order across every page, so you can poll a product for new reviews; without `sort` Trendyol returns its own 'recommended' order, which is not chronological. Every page carries the real totals (average rating, how many people rated, how many wrote a comment, star breakdown, total_pages for your page_size) so you know when to stop. Reviewer names are published by Trendyol already masked. | Catalog enrichment teams call product/reviews to get paginated customer reviews by `content_id` OR `url`, in the order you ask for. | content_id, url, page, page_size, sort |
| product/similar | Trendyol's own 'Benzer Ürünler' recommendation rail for a product, by `content_id` OR `url` — the products Trendyol itself puts in front of a shopper looking at this one, so you get its live view of the competitive set (same category, comparable specs) rather than a keyword guess. Each row is a full product card (price, list price, rating, review count, seller, stock) usable directly for price benchmarking, or feed results[].content_id straight into product/detail. Paginated 0-BASED; meta.pagination.has_more tells you when the rail is exhausted. | Retail analysts call product/similar to get trendyol's own 'Benzer Ürünler' recommendation rail for a product, by `content_id` OR `url`. | content_id, url, page, size |
| search | Search Trendyol products by keyword with optional sort & price/brand filters → results[]{content_id,title,brand,url,price,list_price,rating,comment_count,category,image,free_cargo}. Resolves content_ids for product/detail. Paginated+deduped (page 1-based, max_pages ≤10). meta.total_count = full result-set size. | Pricing teams call search to search Trendyol products by keyword with optional sort & price/brand filters → results[]{cont…. | query, page, max_pages, sort, price, ... |
| search/filters | Discover the available filter facets for a `query` OR `category_id` before filtering — brands (with ids for the `brand` param), price ranges, product ratings, and category-specific attributes (RAM, screen size, color, …) each with live result counts. | Marketplace operators call search/filters to discover the available filter facets for a `query` OR `category_id` before filtering. | query, category_id |
| category/products | Browse a Trendyol category by id — best-sellers, newest, or price/brand-filtered product lists. Same card shape as search. Get a category_id from a product's category.id (product/detail) or the '-c<id>' in a category URL. | Catalog enrichment teams call category/products to get browse a Trendyol category by id. | category_id, url, page, max_pages, sort, ... |
| seller/products | All products of a Trendyol seller/store (paginated, sortable, filterable) by seller_id. Same card shape as search. meta.total_count = the seller's full catalog size. | Retail analysts call seller/products to get all products of a Trendyol seller/store (paginated, sortable, filterable) by seller_id. | seller_id, url, page, max_pages, sort, ... |
| seller/info | Trendyol seller/store profile by seller_id: store name, legal entity (official_name), total product_count, seller score, and trust badges. | Pricing teams call seller/info to get trendyol seller/store profile by seller_id. | seller_id, url |
| seller/categories | Trendyol's seller category tree, as used for product uploads: every category with its id, full path and whether it is a leaf (products are uploaded to leaf categories). Filter with `query`; take the id to seller/category-attributes for that category's upload template. | Marketplace operators call seller/categories to get trendyol's seller category tree, as used for product uploads. | query, leaf_only, limit |
| seller/category-attributes | The product upload template of one Trendyol category: every attribute a listing in it takes, whether it is required, whether a custom value is allowed, whether it creates variants (varianter) or splits listings (slicer), and the allowed values. | Catalog enrichment teams call seller/category-attributes to get the product upload template of one Trendyol category. | category_id |
| suggest | keyword suggestions for a query (ops; search-fallback if suggest API blocked) | Retail analysts call suggest to get keyword suggestions for a query (ops; search-fallback if suggest API blocked). | query |
Call search from your stack
curl -X POST https://api.reefapi.com/trendyol/v1/search \
-H "x-api-key: $REEF_KEY" \
-H "content-type: application/json" \
-d '{"query":"iphone"}'import requests
r = requests.post(
"https://api.reefapi.com/trendyol/v1/search",
headers={"x-api-key": REEF_KEY},
json={
"query": "iphone"
},
)
print(r.json()["data"])const res = await fetch("https://api.reefapi.com/trendyol/v1/search", {
method: "POST",
headers: {
"x-api-key": process.env.REEF_KEY,
"content-type": "application/json",
},
body: JSON.stringify({
"query": "iphone"
}),
});
const { ok, data, meta, error } = await res.json();Ask your MCP-connected assistant: call reefapi.trendyol.search with {"query":"iphone"}.Who uses this API and why
- Pricing teams call search and product/detail to track Trendyol prices, discounts and coupons.
- Catalog-enrichment tools use product/detail and reviews to fill listings with specs and ratings.
- Seller-intelligence products use seller/products to audit a competitor's full Trendyol catalog.
Questions developers ask before integrating
Why does price.current say "49.999 TL" for a phone that costs 50,000 lira?
Because it is Turkish number formatting, where the dot separates thousands. "49.999 TL" is 49,999 TRY, not 49.999. The same response also returns price.current_value as the plain number 49999, and search rows pair price "74,90 TL" with price_value 74.9, a comma for decimals in the same API. Parse the *_value fields and treat the formatted strings as display text only.
What is the difference between content_id and product_group_id?
content_id identifies one buyable product page, the digits after -p- in the URL, and it is the only id the product actions accept. product_group_id identifies the group its colour and storage variants belong to. The iPhone 15 128 GB Mavi returned content_id 762254881 and product_group_id 580170099 on 2026-08-27. Sending a product_group_id where a content_id is expected will not resolve, so store both if you are tracking variants.
Why does rating.comment_count not match rating.total_count?
They count different things. comment_count is written reviews; total_count is star ratings, including people who rated without typing anything. The same product returned comment_count 939 against total_count 1986. The reviews action names them more clearly as total_comment_count and total_rating_count, and it also returns total_image_review_count, 259 there, for reviews with photos. Only the written reviews are pageable.
Why do product/detail and product/reviews report slightly different ratings?
They read different surfaces and round differently. On 2026-08-27 detail returned rating.average 4.644511581067472 while reviews returned summary.average_rating 4.6, and the rating totals were one apart, 1986 against 1987, because the counts moved between the two calls. Use detail's average when you need precision, and treat a one-unit drift in the counts as normal for a live marketplace rather than a parsing bug.
Reviews come back empty on page 1. What am I doing wrong?
product/reviews is 0-based: the first page is page 0, then 1, 2 and so on. Everything else in this engine (search, category/products, seller/products) is 1-based, which is what makes it easy to miss. page_size runs 1 to 50 and anything larger is capped at 50, and summary.total_pages is recomputed for whatever page_size you sent, so it is not a constant for the product.
Why is rating null instead of 0 on some search results?
null means nobody has rated the product yet, which is not the same as a rating of zero. In a live laptop search, rows with no feedback returned rating null and comment_count null, while a rated row returned 4.888888888888889 and 9. list_price behaves the same way: it is null with discount_pct 0 when the seller is not showing a struck-through price, rather than repeating the current price. Check for null before doing arithmetic on any of these.
What is in other_offers, and why are the seller names null there?
other_offers lists the competing sellers for the same product. The iPhone 15 above had 17 of them, from 51.999 TL upward. Each entry carries price, price_value and the campaign text attached to that offer, but seller_id, seller_name and seller_score came back null across all of them on 2026-08-27. The main seller block is fully populated: id 968, name "Trendyol", official_name "DSM GRUP DANIŞMANLIK İLETİŞİM VE SATIŞ TİCARET ANONİM ŞİRKETİ", score 9.3. Use other_offers for the price spread and seller/info when you need the merchant behind a specific id.
How do I filter a search by brand?
The brand parameter takes Trendyol's numeric brand id, not the brand name. Sending "LENOVO" does nothing, because the parameter ignores an invalid value and silently drops the filter. Call search/filters with the same query first and read the "Marka" facet: for "laptop" on 2026-08-27 it returned LENOVO 102323 with 28,761 products, HP 101849 with 18,345, ASUS 101606 with 12,959 and MSI 107655 with 7,561. The same call also returns category-specific facets with live counts, such as processor type and graphics card.
What is the Trendyol API?
Trendyol API is a ReefAPI endpoint group for turkish marketplace products, prices and reviews. It returns live JSON through POST requests under /trendyol/v1.
Is the Trendyol API free to try?
Yes. ReefAPI starts with 1,000 free credits, no card required. Trendyol calls use the same shared credit balance as every other ReefAPI engine.
Do I need a Trendyol login or account?
No login to Trendyol is needed for the API response. You call ReefAPI with your x-api-key header, and the playground can run live examples before you create a production key.
How fresh is the Trendyol data?
The page example is captured from a live product/offers call, and production requests fetch live data through ReefAPI rather than a static sample.
How many credits does the Trendyol API use?
Trendyol actions currently cost 1-2 credits per successful call. Failed or blocked calls are free. All APIs draw from one credit pool.
Can I call Trendyol from an AI assistant or MCP client?
Yes. Connect ReefAPI once through MCP and your assistant can call trendyol actions with the same key, credit pool and JSON envelope used by normal REST requests.