Flipkart Scraper — Prices, Discounts & Price Drops

Give it search words or Flipkart list URLs and get one clean row per product: price, MRP, discount %, rating, ratings count, stock, F-Assured and sponsored flags, image and link, in the order Flipkart lists them. $1 per 1,000 products, up to 1,000 products per search, no login, no API key.

Run it again — or put it on a daily schedule — and every row also tells you what changed since the last run: the previous price and the change, the lowest price seen so far, rank moves and new products.

What you get

One row per product (type: "product"):

Field	What it is
search, sort, rank, page	which search found it, the order used, its position in that order (sponsored slots count) and the page it came from
productId, url, title, brand, imageUrl	the product. brand is filled when Flipkart's card carries it (see Notes)
price, mrp, discountPct, currency	selling price in whole rupees, the struck-through list price (MRP), the discount % as Flipkart shows it (it rounds down), INR
rating, ratingsCount, reviewsCount	average stars, number of ratings, number of written reviews. Counts only: no reviewer names or review texts. rating is empty (not 0) for products with no ratings yet
inStock	true when Flipkart lets you buy it, false when it is out of stock, unavailable or coming soon
isAssured, isSponsored	has the F-Assured badge (null when the page Flipkart sent carried no badge data for that card, see Notes); is shown as a paid placement in the list
keySpecs, tags	headline specs where Flipkart lists them (mostly electronics and appliances) and the card's tags such as "Bank Offer", "Only 3 left" or "Lowest price in the year"
previousPrice, priceChange, priceChangePct, lowestPriceSeen	price in your previous run, the change (negative = cheaper), the change in %, and the lowest price this Actor has seen for the product
previousRank, rankChange, isNew	rank in your previous run, rankChange (positive = moved up), and true when the product was not in the earlier results
observedAt	when it was read (ISO time)

On the first run the change fields (previousPrice, priceChange, priceChangePct, previousRank, rankChange, isNew) are null and lowestPriceSeen is the current price.

One free summary row per search (type: "summary"): productsReturned, pagesRead, totalResultsOnFlipkart (the count Flipkart reports), sponsoredProducts, duplicatesSkipped, pagesRefetched, newProducts, priceDrops, priceRises, droppedOut (products from the last run that are no longer in the results; null when this run did not read as far as the last one), stoppedBecause (limit, end-of-results, no-results or error), previousRunAt, searchUrl.

Run it again and again

• On a schedule — save your input as a task and add a Schedule: the same searches every day. Send each run to Google Sheets, Slack or email with an integration on the task.
• Watch for changes — Keep the same input on the schedule. Each run is its own dataset with the date it ran, so the series builds up run by run and you can compare one day with the last.
• In bulk — Put every search in Searches in one run.

Use it for

• Price-drop and deal tracking — schedule it daily and keep the rows where priceChange is below 0, or where price equals lowestPriceSeen
• New arrivals — sort by newest and keep isNew: true
• Category and competitor research — discount depth, F-Assured share, ratings counts and brands for any search or category list
• Reselling and sourcing — compare selling price with MRP across a whole category

Input

{
  "searches": ["iphone 15", "running shoes", "https://www.flipkart.com/mobiles/pr?sid=tyy%2C4io"],
  "maxProductsPerSearch": 100,
  "sort": "price_low_to_high"
}

• searches — words or phrases, or Flipkart list URLs: a search URL (flipkart.com/search?q=…) or a category list (flipkart.com/…/pr?sid=…). Product page URLs are refused with a message.
• maxProductsPerSearch — 1 to 1000, default 100.
• sort — relevance (default), popularity, price_low_to_high, price_high_to_low or newest.
• readProductPages — off by default and rarely needed. Turn it on to open the product page of every product that still has no rating data after the list was read (see Notes).

The previous run's prices and ranks are kept per search and sort in a named key-value store in your Apify account, so the second run of the same search already shows changes. The store also holds the anonymous visitor cookie Flipkart gave the Actor for that search (see Notes). Schedule it (Apify → Schedules) for a daily history.

Pricing

$1 per 1,000 products, plus $0.01 per run. Error rows and the per-search summary rows are free. The $0.01 run-start fee is the one charge every run pays, even one that returns nothing.

Notes

All numbers below were measured on 9 October 2026 (JST) with this Actor, 200 products per search unless stated.

• Public data only: what any visitor sees on flipkart.com without signing in. The Actor does not set a delivery PIN code. It never returns reviewer names or review texts. Flipkart is India-only, so there is no country or language setting; prices are in rupees.
• Stock depends on location for groceries. For "fresh vegetables" 199 of 200 results said "Currently unavailable" (inStock: false) with no PIN set. Flipkart also lists out-of-stock products in ordinary searches: 127 of the first 200 "iphone 15" results were unavailable, nearly all on pages 4 to 10.
• Flipkart sends some visitors a thinner list. The same page fetched 8 times without cookies gave full cards twice and thin cards six times (T-shirt page 1). Thin cards have no rating, stock, F-Assured, brand or key specs. The Actor keeps Flipkart's anonymous visitor cookie, and when a page comes back thin it asks again with a new cookie, up to 6 times per search (pagesRefetched in the summary says how often). Result: a rating on 182 of 200 T-shirts, 184 watches, 189 shoes, 195 books and 190 sofas. Without this, repeated runs gave a rating on 5 to 45 of 200 T-shirts and 0 to 88 of 200 watches. If every try is thin, rating, ratingsCount, reviewsCount, keySpecs and isAssured stay empty (isAssured is null, not false) and inStock comes from whether Flipkart lets you buy the product. Then readProductPages can fill rating, counts and stock from the product pages (about one extra second per product).
• The order is more stable when the same visitor cookie is used, so the Actor saves the cookie with each search's previous results and sends it again next time. For "wireless earbuds" in relevance order, 77 of 80 results kept the same rank 20 seconds later with the same cookie and 13 of 80 with a new one. It is still not exact: in two repeat runs 15 seconds apart (60 products per search, relevance order, three searches) 26 to 55 products per search kept their exact rank, and 1 to 4 per search were flagged isNew within those seconds. Read isNew as "was not in the list last time", not "just launched". priceChange compares the same product and is not affected. Price order shuffles products with equal prices: for "wireless earbuds" sorted by price only 29 of 80 kept their rank 20 seconds later, though 79 of 80 were the same products.
• About 1,000 products per search at most. "mobile cover" sorted by newest: 25 pages of 40 cards = 1,000 cards (932 different products), page 26 empty. "laptop" sorted by price: 41 pages of 24 = 984 cards (913 different products).
• Flipkart repeats products across pages. In relevance order the same sponsored placements come back on every page: for "watch", pages 2 to 6 each repeated 14 to 25 of their 40 cards, and 14 to 16 of those were sponsored. The Actor keeps each product once, at its first position (duplicatesSkipped counts the repeats: about 30 per 200 products for "iphone 15" and 110 to 175 for most other searches I tried), so reading 200 products took 7 to 16 pages, and a search may return fewer products than you asked for.
• isNew and droppedOut compare with your previous run of the same search and sort. If you raise maxProductsPerSearch, products beyond the earlier range get isNew: null (unknown) instead of true; a run with a lower limit replaces the saved range with its own. A run that stops early with an error does not replace the saved results unless it read at least as many products as the earlier one.
• Some fields are only filled in some cases. mrp is empty when Flipkart shows no struck-through price (13 or 14 of 200 "iphone 15" results had one, 6 of 200 "fresh vegetables"; 190 or more of 200 for laptops, washing machines and books). brand is empty when the card has no brand field: filled for all 200 "iphone 15", "mobile cover", "washing machine", "tshirt", "watch" and "books", none of 200 "laptop" in my last run (0 to 64 in four runs), 3 of 200 "fresh vegetables"; the title normally starts with the brand. keySpecs was filled for all 200 results of "laptop", "washing machine", "iphone 15", "tshirt", "watch" and "books", for none of "fresh vegetables", and for 0 and 40 of 200 "mobile cover" in two runs. tags is empty when the card has none.
• isSponsored marks the list's paid placements (none in "iphone 15", 21 to 32 of 200 T-shirts and 18 to 32 of 200 watches across my runs). They keep their position in rank.
• Flipkart widens searches it does not understand: the made-up word "asdfqwerzxcv" still returned over 4,000 results, so check title when a search is vague.
• If Flipkart answers with its bot-check page, the Actor stops, returns a free error row for the search and the ones it did not read, and does not try to get around it. Five searches of 200 products (1,000 rows, 45 pages) took 67 seconds.

To track Shopify stores the same way — prices, new products and sold-out items — see Shopify Store Products & Price Tracker.

For AI agents

{"searches": ["iphone 15"], "maxProductsPerSearch": 5} returns the first 5 Flipkart products for "iphone 15" with price, MRP, discount, rating and stock, plus one summary row. Rows have type: product, summary or error. Only product rows are charged.

**Something broke or looks wrong?** Open an issue on this Actor's Issues tab and I'll look at it. **Found it useful?** A short review on the Store page helps other people find it.