Yandex Maps Business Scraper

Build clean, ready-to-work local business lists from Yandex Maps. Give it a search term and a city and you get back every place the map publishes for that search — the business name, phone number, website, social profiles, full street address, exact coordinates, star rating, review count and opening hours — as one tidy table you can hand to a sales team, load into a CRM or open in a spreadsheet. It is built for people selling into Russia, Belarus, Kazakhstan and Uzbekistan, where Yandex Maps is the map that local businesses actually keep up to date.

What you can do with it

• Build an outbound list for a whole city. Search "dentist", "fitness club" or any other term and get hundreds of businesses with a phone number and a website attached to each one.
• Sweep a city properly. Draw a rectangle on the map, split it into a grid, and the actor searches every square separately — which reaches places a single city-wide search never shows you.
• Target a neighbourhood. Give a centre point and a radius in kilometres and collect only what sits inside it.
• Study a competitive landscape. Ratings, review counts, chain membership and opening hours for every player in a category, in one table.
• Enrich what you already have. Match on name and address to fill in missing phone numbers, websites and social profiles.
• Track a market over time. Schedule it weekly and watch new openings, closures and rating movements.

What you get

One row per business. An abridged real record:

{
  "orgId": "1287729374",
  "name": "Столица",
  "primaryCategory": "Стоматологическая клиника",
  "categories": ["Стоматологическая клиника", "стоматологическая поликлиника", "детская стоматология"],
  "categoryIds": ["184106132", "184106130", "64561926512"],
  "phone": "+74951261186",
  "phones": [{ "number": "+74951261186", "formatted": "+7 (495) 126-11-86", "type": "phone", "extension": null, "note": null }],
  "website": "https:\/\/stolica-dent.ru\/",
  "websiteDomain": "stolica-dent.ru",
  "socialLinks": [{ "platform": "telegram", "url": "https:\/\/t.me\/stolicadent_msk", "handle": "@stolicadent_msk" }],
  "address": "просп. Мира, 12, стр. 3",
  "fullAddress": "Москва, проспект Мира, 12, стр. 3",
  "postalCode": "129090",
  "country": "Россия",
  "latitude": 55.77406,
  "longitude": 37.632557,
  "nearestMetroStation": "Сухаревская",
  "nearestMetroDistanceMeters": 141,
  "rating": 5,
  "ratingCount": 1139,
  "reviewCount": 735,
  "openingHoursText": "Mon 00:00-24:00; Tue 00:00-24:00; Wed 00:00-24:00; Thu 00:00-24:00; Fri 00:00-24:00; Sat 00:00-24:00; Sun 00:00-24:00",
  "isVerified": true,
  "isChain": false,
  "chainName": null,
  "averageBill": null,
  "photoCount": 151,
  "registeredCompanyName": "ОБЩЕСТВО С ОГРАНИЧЕННОЙ ОТВЕТСТВЕННОСТЬЮ \"СТОЛИЦА\"",
  "registeredCompanyTaxId": "7702427452",
  "orgUrl": "https:\/\/yandex.com\/maps\/org\/stolica\/1287729374\/",
  "searchTerm": "стоматология",
  "regionId": 213,
  "regionName": "Москва",
  "position": 2,
  "collectedAt": "2026-08-17T10:00:00.000Z"
}

Input reference

Field	Type	Default	What it does
searchTerms	list of text	—	What to look for, one term per line. Terms in the local language usually return more places than their English equivalents.
categoryIds	list of text	—	Search whole categories instead of typing a term. Every result carries its own category codes, so one exploratory run gives you the codes for everything you want to sweep next.
regionId	choice	Moscow	The city the search runs in. 21 cities across Russia, Belarus, Kazakhstan and Uzbekistan.
language	choice	Local spelling	Whether names, addresses and categories come back in local spelling (Ригла, ул. Арбат) or romanised (Rigla, Arbat Street).
maxResultsPerQuery	number	200	The most businesses to collect for each term in each area. Maximum 1,000.
boundingBox	text	—	Restrict the search to a rectangle, written as two opposite corners: longitude,latitude~longitude,latitude.
centerLatitude, centerLongitude, radiusKm	numbers	—	Search around a point instead of a rectangle. Fill in all three together.
tileGrid	number	1	Splits the search area into a grid and searches each square separately. 2 gives 4 squares, 3 gives 9, 4 gives 16.
onlyWithPhone	yes/no	no	Skip places that publish no phone number.
onlyWithWebsite	yes/no	no	Skip places that publish no website.
minRating	number	—	Keep only places rated at least this highly. Unrated places are left out when this is set.
includePopularTimes	yes/no	no	Adds an hour-by-hour busyness profile per weekday, where the source publishes one.
parallelSearches	number	3	How many searches to run at a time.

Output fields

Field	What it holds
orgId	The place's own identifier on the map — stable, and the right key to de-duplicate on.
name, shortName	Business name as published.
primaryCategory, categories, categoryIds	What kind of business it is, and the codes you can search by.
phone	The first published number, ready to dial, in international form.
phones	Every published number, with its display form, any in-house extension and any label the business attached to it.
website, websites, websiteDomain	Published web addresses, with the map's own tracking decoration stripped off.
socialLinks	Social profiles with platform, address and handle.
address, fullAddress, postalCode, country	Street address in short and full form.
latitude, longitude	Exact position.
nearestMetroStation, nearestMetroDistanceMeters	Closest metro station and how far away it is in metres.
rating, ratingCount, reviewCount	Star rating, how many people rated it, how many wrote a review.
openingHours	Opening times per weekday as structured intervals, in 24-hour clock.
openingHoursText	The same schedule on one line, always in the same fixed format.
timeZone	The place's own time zone, written as UTC+03:00, so opening times can be compared across cities.
isVerified	Whether the owner has claimed and confirmed the listing.
isChain, chainName	Whether the place belongs to a chain, and which.
averageBill	Typical spend, where the business publishes one.
photoCount	How many photos the listing carries.
registeredCompanyName, registeredCompanyTaxId	The legal entity behind the listing and its tax number, taken from the advertising disclosure a promoted listing has to carry. Present on promoted listings only.
orgUrl	Direct link to the listing.
popularTimes	Hour-by-hour busyness per weekday, when you switch it on.
searchTerm, regionId, regionName, requestedRegionId, searchArea, position	What was searched, where, and where this place came in the results.
collectedAt	When the record was collected.

Anything the business has not published comes back empty rather than guessed. A place with no website has no website in your table — never a blank string standing in for one, and never a value borrowed from a neighbouring record.

Pricing

You pay for what you get, with no monthly fee.

What you pay for	When you pay it	Price
Business found	For each business saved to your results	$6.00 per 1,000
Area scanned	Once for each search term in each area, when that search finishes — including when it turns out to contain nothing	$0.02 each

A search that could not be completed is never charged.

A worked example. One search term in one city, collecting 200 businesses: 200 × $0.006 + one area at $0.02 = $1.22. A full city sweep with 4 search terms over a 3 × 3 grid — 36 searches — collecting 2,400 businesses in total: 2,400 × $0.006 + 36 × $0.02 = $15.12.

The second charge exists because a search is real work whether it returns 250 businesses or 2, and it keeps the per-business price low for the runs that return a lot.

Limits & what this actor cannot do

• Results are a snapshot at the moment of collection, and the source's own results change continuously.
• The actor reports what the source publishes to the public. It does not sign in, does not see anything behind a login, and cannot reveal a detail the source keeps private.
• Contact details are reported only where the source publishes them. Measured across 124 businesses in 5 cities and 6 lines of business: a phone number on all of them, a website on 94%, social profiles on 92%, opening hours on 99% and a published rating on about 97% — deeper result pages carry more places that nobody has rated yet, and those come back empty rather than as a zero. The remainder legitimately publish nothing, and those come back empty rather than guessed.
• There are no email addresses on this source. Nothing in the listing carries one, and the actor will never invent one.
• The source publishes a limited number of places for any one search in any one area — around 1,000 positions, with some places repeated, so a broad term in a single area usually settles at several hundred distinct businesses. Splitting the area into a grid is how you go further, and it genuinely reaches different places rather than re-cutting the same list.
• The count of matches the map shows on screen is an estimate that moves as you look deeper, so the actor never reports it. What you get is a count of what was actually collected.
• Results differ by city and by language; every row is stamped with the city the source reported for it, which is not always the city that was asked for.
• Ratings and review counts are as published at collection time and can lag behind the reviews themselves.
• A registered company name and tax number come from the advertising disclosure that a promoted listing has to carry, so they arrive only on promoted listings and how many rows have them depends entirely on how heavily the category is advertised. Measured live: 60% of a Moscow dentistry list and 50% of a central-Moscow restaurant list, and expect far less in a category nobody advertises in. The rest come back empty rather than guessed.
• A search the source refuses to complete is reported as a failed search, never as "no results found". The run summary counts searches that produced businesses, searches that were genuinely empty, and searches that failed, as three separate figures.
• Speed depends on the size of the job and on the source's own response times; no fixed throughput is promised.
• The source's terms govern automated access. You are responsible for using the data lawfully and in line with those terms, and for handling any personal data in line with applicable privacy law — including GDPR and Russian personal data law where they apply.

FAQ

Do I need an account on the source site? No. No account and no login are needed — just run it.

Does it need my login or password? No. It never asks for credentials of any kind and never signs in anywhere.

Can I schedule it? Yes. Schedule it daily, weekly or monthly, and each run produces a fresh table.

Why do I get fewer businesses than the map says exist? Because the source publishes only part of its list for any one search in any one area, and the figure shown on screen is an estimate that changes as you look deeper. Draw a rectangle and set a grid — each square is searched separately and reaches places the single city-wide search never shows.

Are email addresses included? No. This source does not publish them, so no honest tool can return them. What you do get is a phone number on virtually every business and a website on most, which is what an outbound team actually works from.

Which language should I choose? Local spelling matches what a Russian-speaking sales team will see and search for. Romanised spelling is easier to load into an English-language CRM and easier to read at a glance. The businesses returned are the same either way.

Can I search a whole country at once? No — searches run city by city, because that is the level at which the source returns dependable results. The 21 cities offered were each checked against live results, so the places you get back really are from the city you picked.