Swiggy Scraper: Restaurants, Ratings & Cuisines

Scrape Swiggy restaurant listings by Indian city — name, rating, cuisines, cost for two, delivery time, offers and locality. No login, pay per restaurant. Food-delivery market research and competitor tracking. Use it as an MCP server in Claude, ChatGPT & AI agents.

⚡ Part of a 74-scraper suite with 8,800+ runs · no login, no cookies, no ban risk.

💸 You are only charged for delivered results — empty searches, failed pages and duplicates are never billed.

What it does

Swiggy is India's largest food-delivery platform. Its restaurant listings are geo-indexed — what you see depends on the coordinates you ask from — so a single request only ever returns the handful of restaurants serving one point.

This actor handles that properly: it samples several coordinates across a city and merges the results, deduplicating by restaurant id, so you get genuine city-level coverage instead of one neighbourhood. Each record carries the restaurant id and name, city, area and locality, cuisines array, rating and rating count, cost for two, delivery time in minutes, distance, any active offer and a canonical URL.

Input

Field	Type	Required	Default	Description
city	string	no	"bangalore"	Indian city to scrape.
latitude	string	no	—	Override the city centroid with an exact latitude.
longitude	string	no	—	Override the city centroid with an exact longitude.
maxResults	integer	no	10	Maximum restaurants to return.

Example input

{
  "city": "bangalore",
  "maxResults": 10
}

Output

One record per restaurant. Field list and example below are taken from a real run, not a mock-up.

Field	Type	Example
restaurant_id	string	"10575"
name	string	"Pizza Hut"
city	string	"bangalore"
area	string	"Central Bangalore"
locality	string	"Richmond Town"
cuisines	array	["Pizzas"]
rating	number	4.3
rating_count	string	"12K+"
cost_for_two	string	"₹600 for two"
delivery_time_mins	number	27
distance	string	"2.7 km"
offer	string	"ITEMS"
image_id	string	"RX_THUMBNAIL/IMAGES/VENDOR/2026/6/15/6f892…
url	string	"https://www.swiggy.com/restaurants/pizza-h…
scraped_at	string	"2026-07-27T08:45:07.223Z"

Example output

{
  "restaurant_id": "10575",
  "name": "Pizza Hut",
  "city": "bangalore",
  "area": "Central Bangalore",
  "locality": "Richmond Town",
  "cuisines": [
    "Pizzas"
  ],
  "rating": 4.3,
  "rating_count": "12K+",
  "cost_for_two": "₹600 for two",
  "delivery_time_mins": 27,
  "distance": "2.7 km",
  "offer": "ITEMS",
  "image_id": "RX_THUMBNAIL/IMAGES/VENDOR/2026/6/15/6f892224-627c-4702-89e5-430fb3637e58_10575.JPG",
  "url": "https://www.swiggy.com/restaurants/pizza-hut-10575",
  "scraped_at": "2026-07-27T08:45:07.223Z"
}

Use cases

• Restaurant market mapping — see who operates in a city, in which cuisines, at what price point.
• Competitive rating benchmarking — compare a brand's rating and review volume against every competitor in the same locality.
• Cloud-kitchen discovery — brands appearing at many coordinates with identical menus are typically cloud kitchens.
• Pricing analysis — cost_for_two and the offer field together show real positioning versus list positioning.
• Delivery-logistics research — delivery_time_mins and distance across a coordinate grid reveal effective service radii.
• Site selection — cuisine density by locality highlights under-served pockets.

FAQ

Why does the actor sample multiple coordinates?

Because Swiggy's listing surface is geo-indexed and caps at roughly 8-28 restaurants per coordinate. Requesting one point per city would silently return a fraction of the market and look like complete coverage. Sampling several points across the city and deduplicating is what makes city-level results real.

Can I target a specific neighbourhood?

Yes. Leave latitude and longitude blank to sample the whole city, or set them to pin the search to one point.

Is menu or dish-level data included?

No. This actor returns restaurant-level records. Menu extraction is a separate, much heavier job and is not claimed here.

Is rating_count a number?

It is Swiggy's own display string, such as '12K+', because that is the precision Swiggy publishes. Converting it would invent accuracy that the source does not have.

Billing

Pay per event: one restaurant-scraped event per delivered record, charged only after the record is validated and stored. Empty results, blocked pages and duplicates are never charged. Store discounts apply automatically on paid Apify plans.

Use in Claude, ChatGPT & any MCP agent

This actor runs as an MCP server, so an AI agent can call it directly:

https://mcp.apify.com/?tools=themineworks/swiggy-scraper

from apify_client import ApifyClient
client = ApifyClient("YOUR_APIFY_TOKEN")
run = client.actor("themineworks/swiggy-scraper").call(run_input={"city": "bangalore", "maxResults": 10})
for item in client.dataset(run["defaultDatasetId"]).iterate_items():
    print(item)

Legal and compliance

This actor collects only publicly available information — the same data any visitor can see without logging in. It does not access private or authenticated areas, does not attempt to bypass paywalls, and collects no personal data beyond what the source already publishes openly. You are responsible for using the output in line with the source site's terms and with the data protection law that applies to you.

MIT © The Mine Works