Blinkit Scraper: Prices, Inventory and Dark Stores

Blinkit Scraper collects local product prices, MRP, availability and orderable inventory across Indian delivery locations. Compare the shelf served by each dark store in one dataset.

Use it to monitor an FMCG brand across cities, compare selling prices against maximum retail prices, or record whether a product is available at a particular delivery point. A national product list cannot explain local availability. This actor keeps each product attached to the serving store and the location you requested, so your analysis can distinguish a product missing in one neighborhood from a product available in another.

Blinkit Scraper input

Provide search terms in queries, category identifiers in category_ids, or both. Search terms run independently at every location. The category format is l0 or l0/l1; the Milk category is 14/922. Category identifiers follow Blinkit's category URLs. The actor accepts up to 20 terms and 20 categories per run.

Add one or more entries to locations. Each entry can be a coordinate object with numeric lat and lon, a six-digit Indian pincode, or address text. You can mix those forms in the same list. Pincodes and addresses are resolved using the location information returned by Blinkit. Where a location cannot be resolved or served, the actor reports that limitation instead of assigning an approximate store.

This input uses the delivery point from the recorded Gurgaon search example:

{
  "queries": ["amul milk"],
  "locations": [{"lat": 28.4595, "lon": 77.0266}],
  "max_items": 15
}

max_items limits delivered product offers per location and per search term or category. It defaults to 48 and accepts values from 1 to 500. Product variants count as separate offers. include_out_of_stock defaults to true; turn it off to keep only currently available offers. include_stores also defaults to true and adds fulfillment location context.

For existing integrations, searchQueries and searchTerms are supported aliases for queries. productsLimit and maxResults alias max_items. For search terms, a supplied main field takes precedence. If several limit fields are supplied, the smallest limit wins. A single delivery location can also be specified with latitude and longitude.

Output: offers and fulfillment locations

The dataset contains two row types identified by both record_type and _type. An offer row represents a product at a serving dark store for one requested delivery point and search combination. A fulfillment_location row connects that dark store to the requested delivery location. Join them using location_id and location_index. The latter is the zero-based index of your input location, so two delivery points remain distinct when the same store serves both.

Offer rows include product_id, title, brand, unit, price, mrp, discount_percent, inventory, availability, product_state, image_url, url, merchant_id and location_id. Prices use Indian rupees and currency is INR. country is IN. Search context includes query, category_id and the one-based position within that search combination. Every row carries source and a UTC fetched_at timestamp.

The recorded first Gurgaon offer is Milky Mist 25 g High Protein Milk, product 1453864, served by merchant 43483. Its selling price is ₹79 against an MRP of ₹99, with a rounded discount of 20%, a 250 ml unit and orderable inventory of 2. These are observations from the recorded response, not a promise that the current shelf will have identical prices.

Fulfillment rows include platform, location_id, delivery_latitude, delivery_longitude, the delivery pincode when available, and offers_count. City, locality, state, formatted address and delivery type are included when returned by the source. The coordinates describe your delivery point. They do not locate the warehouse building. Store rows appear after that delivery point has finished collecting its offers.

Use cases for a Blinkit price tracker

• FMCG brand monitoring: check the selling price, MRP and discount of your range in several cities and see where a product is out of stock.
• Price tracking: schedule the actor daily and compare price and mrp per product_id over time to build a Blinkit price history.
• Availability analysis: use inventory, availability and merchant_id to map which dark stores carry a product around a delivery point.
• Market research: compare assortment and discounting between neighborhoods and cities for quick-commerce analysis.

Blinkit Scraper compared with other options

Blinkit offers no public API for shelf data, so searches for a Blinkit API usually end in a scraper. Many Blinkit tools return one flat product list for a single place. This actor takes several locations in one run and returns, next to the offers, a fulfillment_location row per delivery point that names the serving store. That makes dark-store coverage a direct output instead of something you reconstruct. Existing input aliases stay supported, so you can move over from similar tools without rewriting your input.

Inventory, variants and delivery estimates

Inventory is the quantity currently orderable through the product card, capped at 50. A value of 50 does not establish the exact stock level inside the warehouse. Availability follows Blinkit's product state: InStock or OutOfStock. Unavailable variants can still appear when out-of-stock inclusion is enabled.

Variants have their own product identifiers. Repeated copies of the same card within one search combination are removed. A product matching two different terms can appear once for each term, with the corresponding context preserved. Store offer counts include those delivered search-context rows.

eta_minutes is included when Blinkit displays a recognizable delivery estimate and is null when unknown. Store-level estimates use the most common estimate among that store's delivered offers. They describe the observed delivery promise and should not be treated as a service guarantee.

Pagination and incomplete results

Large requests continue automatically on Apify. Direct API users receive a cursor and repeat the same input with that token until it becomes null. Keep the search terms, locations and options unchanged while continuing. Positions, duplicate detection and store offer counts carry across those requests.

partial and warnings explain unfinished work, empty search combinations or unavailable locations. Other usable locations can still produce results. If every requested combination returns no products, the run has no offer rows. The actor does not substitute products from unrelated locations or invent missing identifiers.

Pricing and trial limits

Product billing counts delivered offer rows only. Fulfillment location context does not generate an offer event. A separate actor start event may apply according to the current pricing shown on this actor's Pricing tab, including runs without matches. Consult that tab for the active amounts before running a large collection.

Non-paying accounts have a delivery cap of 10 dataset rows per run. Offers are ordered before fulfillment context, so the cap can omit some or all store rows. Only the offers actually delivered are counted for product billing. Your requested limit applies independently to each location and term; adding locations or terms increases the potential paid result count.

Frequently asked questions

Can I compare several cities in one run?

Yes. Supply the delivery coordinates, pincodes or addresses together in locations. Each offer keeps its input location index and serving merchant identifier. Compare matching product identifiers across those locations while retaining the timestamps.

Does inventory show exact warehouse stock?

It shows the orderable amount visible on the product card, up to 50. It does not reveal an exact physical count above that limit. Combine inventory with availability when interpreting a missing or unavailable variant.

Are these coordinates the dark store's address?

They are delivery coordinates. The serving dark store is identified by merchant_id and location_id. The actor does not provide a warehouse coordinate or address when Blinkit does not expose one.

Can I start with a pincode?

Yes. A six-digit Indian pincode can be provided as a location string. The returned city and delivery coordinates reflect the source's location resolution. For precise neighborhood comparisons, provide the exact delivery coordinates instead of relying on a broader postal area.