myntra-scraper

A Python scraper for Myntra, India's largest fashion marketplace. Two jobs:

• Listings and search: any listing or search URL (/men-tshirts, /levis?rawQuery=levis, with sort and filters) → name, brand, price, MRP, discount, rating, rating count, sizes, colour, image and URL. Real pagination, 50 products per page, sponsored products left out.
• Product pages: price, MRP, rating and attributes, plus one row per size and seller with that seller's price and the exact units in stock.

Output as Python dicts, JSON or CSV. One dependency (requests), no browser.

pip install myntra-scraper

Why a plain request fails

Myntra sits behind Akamai Bot Manager (_abck, bm_sz cookies). A script doesn't get an error, it gets nothing:

Pages are fetched through the Scrape.do API with a plain request: no render, no super, 1 credit per page. You need a Scrape.do API token; the free plan includes 1,000 credits a month.

Quick start

from myntra_scraper import MyntraScraper

scraper = MyntraScraper(token="YOUR_SCRAPEDO_TOKEN")  # or set SCRAPEDO_TOKEN

# 3 pages of 50 = 150 products (4 credits: one page visit + three API pages)
rows = scraper.get_listing("https://www.myntra.com/men-tshirts?sort=new", max_pages=3)
for row in rows[:3]:
    print(row["brand"], row["name"], row["price_inr"], row["mrp_inr"], row["discount_pct"])

product = scraper.get_product(rows[0]["url"])
print(product["units_in_stock"], product["sizes_in_stock"], "/", product["sizes_total"], "sizes in stock")
for size in product["sizes"]:
    print(size["size"], size["seller"], size["price_inr"], size["units"])

Command line

export SCRAPEDO_TOKEN=your_token

# 3 pages of a listing, CSV
myntra-scraper listing "https://www.myntra.com/men-tshirts?sort=new" --pages 3 -f csv -o myntra-listing.csv

# product summaries for the first 10 products from that file
myntra-scraper product -i myntra-listing.csv --limit 10 -f csv -o myntra-products.csv

# one row per size and seller
myntra-scraper product -i myntra-listing.csv --limit 10 --sizes -f csv -o myntra-sizes.csv

product reads the url column of the listing CSV, so the commands chain.

Output

Size rows (from the fictional test page in tests/fixtures):

[
  {"product_id": 90000001, "size": "M", "sku_id": 900000012, "available": true,
   "seller": "Example Retail Two", "seller_id": 202, "price_inr": 449, "mrp_inr": 999, "units": 5},
  {"product_id": 90000001, "size": "M", "sku_id": 900000012, "available": true,
   "seller": "Example Retail One", "seller_id": 101, "price_inr": 469, "mrp_inr": 999, "units": 30}
]

Listing columns: product_id, name, brand, price_inr, mrp_inr, discount_pct, rating, rating_count, sizes, gender, article_type, colour, image, url.

Product columns: product_id, name, brand, price_inr, mrp_inr, discount_pct, rating, rating_count, colour, article_type, gender, country_of_origin, sizes_in_stock, sizes_total, units_in_stock, lowest_size_price_inr, highest_size_price_inr, sellers, url (plus attributes and sizes in Python/JSON).

How it works

The data is one JSON object. Myntra renders every page on the server and puts the data in an inline script, window.__myx = {...}. Listings keep their products in searchData.results.products, product pages keep everything in pdpData. The scraper finds the assignment and reads the object with json.JSONDecoder.raw_decode. No CSS selectors.

Sponsored products are separate. A listing page carries the organic results in products and the ads in plaProducts (18 of them on a men's T-shirts page). Only the organic products are returned.

?p=2 is ignored. Ask the HTML for page 2, 3 or 50 and you get the first 50 products again. In our test, page 2 of a listing sorted by newest shared 49 of 50 products with page 1. The site pages through its search API, /gateway/v4/search/<slug>, which:

1. returns 401 without the cookies a page visit sets (at, _abck, bm_sz and others), and
2. ignores p and o as well, unless the request carries a pagination-context header. Every API response returns a new pagination-context header; sending it back moves to the next 50.

Not every visit hands out a usable session: in our tests roughly one visit in four came back with a single cookie, and the API answered 401 to it. The scraper repeats the visit until Myntra's at cookie is there, and gets fresh cookies if the API still says 401.

So get_listing(max_pages=N) visits the listing once for its cookies, then calls the search API N times, passing the cursor from each response to the next request (Scrape.do forwards the cookie and cursor headers with extraHeaders=true). In our test, four API pages returned 200 distinct products.

Listing price vs size price. A listing card has one price. On the product page every size lists its sellers, each with its own price and stock count. On a Levi's T-shirt listed at ₹389, sizes M and XL had a seller at ₹389 while sizes S and L were only available from a second seller at ₹402. The sizes rows keep that split.

Limits

• One location. Prices and stock are what Myntra shows without a delivery pincode.
• Credits. max_pages=1 reads the listing HTML (1 credit). More pages cost one page visit plus one credit per API page.
• Errors. Non-200 responses raise ScrapeError with the status code; from_target tells a Myntra status from a Scrape.do one (a wrong token is a Scrape.do 401). A 200 without Myntra's data (an Akamai page) is treated as a failure. Server errors, 429s and empty pages are retried twice; client errors are not.

Responsible use

The scraper reads listing, product and search pages, which Myntra's robots.txt allows (Allow: /); it does not touch cart, checkout, account or size-chart paths. Product data is not personal data. Read Myntra's terms for your use case and keep request rates modest. This is not legal advice.

Development

git clone https://github.com/bugrahansss/myntra-scraper
cd myntra-scraper
pip install -e .
python -m unittest discover -s tests -v

The tests run offline against fictional fixture pages.

License

MIT

Built and tested with Scrape.do, November 2026.