Zepto Catalogue by Location

Pulls the whole Zepto storefront for an Indian pincode, category by category, with price, MRP, discount, the available quantity the app reports and out-of-stock state. Zepto prices and stocks per dark store, so the same SKU differs between neighbourhoods.

What you get

• Complete category walk: category, subcategory, their slugs and ids, plus position so shelf order survives.
• Price economics per product: price, mrp and discount_percent.
• Stock depth rather than a bare flag: available_quantity and max_quantity alongside out_of_stock.
• A sponsored flag: is_ad marks placements Zepto itself tags as sponsored.
• Store context on every row: store_id and location_label, plus product_id, variant_id, name, brand, pack and image_url for joining.
• Location honesty: a pincode Zepto does not serve is reported as a failed location with the reason, and returns zero rows instead of another city's data.

Input

Give one or more Indian pincodes. Everything else narrows the crawl.

{
  "pincodes": ["122002"],
  "categoryFilter": ["fruits"],
  "delayMs": 400,
  "maxPagesPerCategory": 0,
  "proxyConfiguration": {
    "useApifyProxy": true,
    "apifyProxyGroups": ["RESIDENTIAL"],
    "apifyProxyCountry": "IN"
  }
}

• pincodes: 6-digit Indian pincodes, for example 122002.
• categoryFilter: only categories whose slug or name contains one of these strings, case-insensitive. Empty means all.
• maxPagesPerCategory: cap on pagination. 0 pages until Zepto reports the end.
• delayMs: pause between calls, 400 by default.
• proxyConfiguration: residential country IN is the default and is required in practice.

Output

One row per product, per subcategory, per location. The run above returned 618 rows.

{
  "location_label": "Gurgaon-DLF",
  "store_id": "d8c4ab04-68f0-483f-82be-8b9fb18d17ae",
  "category": "fruits-vegetables",
  "category_name": "Fruits & Vegetables",
  "subcategory_name": "Fresh Vegetables",
  "position": 1,
  "product_id": "d2a42dc1-1b5a-4925-9b66-28a24d2d8830",
  "name": "Onion",
  "brand": "Unbranded",
  "pack": "900 g - 1 kg",
  "price": 62,
  "mrp": 126,
  "discount_percent": 50,
  "available_quantity": 5,
  "out_of_stock": false,
  "max_quantity": 5,
  "is_ad": false,
  "fetched_at": "2026-09-11T08:59:34.879Z"
}

brand reads Unbranded for fresh produce, which is Zepto's own value, not a gap. available_quantity is capped by Zepto's per-order limit, so treat it as a floor on real depth.

Use cases

• Per-store price and discount tracking from price, mrp and discount_percent across pincodes.
• Out-of-stock and low-stock monitoring from out_of_stock and available_quantity.
• Ad-slot share: count is_ad rows by brand within a subcategory.
• Assortment gap analysis by comparing the product_id set between cities.

Notes and limits

• Zepto blocks an IP after roughly 5 automated page loads, answering with an HTTP 202 challenge. A residential Indian proxy is effectively required and is the default.
• Each minted session is pinned to one exit IP. A per-request rotating gateway kills sessions within a few calls, so use sticky sessions.
• An unserved point ends that location cleanly, for example Zepto has no open store for 24.9,81.5 (serviceable=false), with zero rows and failed_locations incremented.
• is_ad comes from Zepto's own sponsored markers. Verified runs in fresh produce returned none, so ad density varies sharply by category.
• Exact coordinates can be supplied instead of pincodes through the advanced locations field.

FAQ

Do I need a proxy? Yes, in practice. Use a residential proxy in India. It is already the default, and sessions must be sticky to one exit IP.

How many results can I get? A full store catalogue. One verified run returned 618 products from a single category filter.

Does it need login or an API key? No. No Zepto account, key or cookie of your own is needed.

How does a pincode pick a store? The pincode is resolved to a point and Zepto binds the session to the dark store serving it. That store's id is stamped on every row as store_id; if no store is open there, the location is skipped with the reason.