🗺 Yandex Maps Places Scraper

The fast, captcha-resilient way to extract every business on Yandex Maps. Names, phones, emails, websites, ratings, reviews, photos, opening hours, menus — and 60+ more fields — straight into your Apify dataset.

---

Why this scraper?

	This actor	Typical Yandex scraper
Cost	$3 / 1,000 places	$4–8 / 1,000
Engine	Direct HTTP (curl_cffi + TLS impersonation)	Headless Chromium
Speed	~150 places/min on a single run	~30–60 places/min
Captcha resilience	Residential proxy + auto session rotation (5 retries)	Often blocks mid-run
Coverage	60+ fields including emails / website phones / reviews / menus	15–25 fields
Regions	6: RU, BY, KZ, TR, AZ, EN	usually RU only
Billing	Pay-per-event — you control what you pay for	Flat rate, you pay for unused

What makes it fast and reliable

• No browser — Yandex server-renders the entire app state into a single JSON payload on every search/place URL. We parse that directly. No Chromium boot, no JS execution, no rendering loop.
• Chrome TLS fingerprint — curl_cffi sends a TLS ClientHello byte-identical to real Chrome, defeating Yandex's JA3-based bot detection.
• Auto proxy pairing — your selected Language automatically picks the matching residential country (RU → RU IPs, BY → BY, TR → TR…). No more cross-region captcha loops.
• Captcha-aware retry — on 403 / showcaptcha / suspicious response we rotate the proxy session, recreate the curl context (clean cookie jar + new exit IP), and retry up to 5 times before giving up.
• Parallel enrichment — reviews and website-contact scrapes run concurrently per place, halving wall-time when you opt into both.
• 15s ceiling per request — broken proxies fail fast, healthy ones return in well under 2s.

---

What you can do with it

Use case	Workflow
🎯 Lead generation	query="dentist", locations=[Moscow, SPB], emails=ON → CSV with names, addresses, websites, emails harvested from each place's website, phones
📊 Local market research	categories=[Pizzeria], locations=[10 cities] → per-city ratings, density, price tiers — perfect for franchise scouting
⭐ Reviews monitoring	organizationIds=[your shops], maxReviews=50, sort=newest → JSON of customer feedback, sortable by date or rating
📚 Directory building	query="carwash", maxItems=2000 → bulk-extract every car wash with full metadata
🕵 Competitor intelligence	one company's placeId → features, opening hours, photos, social links, menu sections, awards

---

Quick start

1. Search queries — type what you'd type in the Yandex Maps search bar (e.g. coffee, auto repair, dentist). One per line.
2. Location — pick one or more cities from the dropdown (~90 major Russian cities listed in English: Moscow, Saint Petersburg, Novosibirsk…) or type your own city name.
3. Number of places to extract — 20 is a good start.
4. Language — sets the Yandex regional domain. Defaults to Russian.
5. (optional) Filter by rating, price tier, category, or "good places only".
6. (optional) Enable reviews, extra photos, or email/phone from each place's website.
7. Save & Start.

💡 No Search queries? Pick a category — the actor will use category names as queries automatically (English + Russian for max coverage).

---

Pricing — Pay only for what you collect

Event	Cost	When it fires
Result	$3.00 / 1,000	Each place pushed to the dataset
Review scraped	$0.50 / 1,000	Each individual review extracted
Image scraped	$0.50 / 1,000	Each extra photo (the main image is free)
Actor start	$0.00005	Once per run
Platform usage (compute, proxy, dataset)	free — included	—

🛡 You set the ceiling. In Run Options set maxTotalChargeUsd (e.g. 5) and the actor will stop before exceeding your budget. Reviews, extra photos, and contact-scraping are off by default — turn them on only if you need them.

Real-world budget examples

Goal	Setup	Cost
500 leads with phones + emails	500 places + contact scraping	$1.50
500 places + 10 reviews each	500 + 5,000 reviews	$4.00
200 restaurants with photos & reviews	200 + 5 photos × 5 reviews	$1.10
2,000 carwashes for a directory	2,000 places, no extras	$6.00

---

Input

Top-level fields

Field	Description
🔎 Search queries	List of search terms. Each query runs against every selected city.
📍 Location	Cities to limit the search. Pre-loaded with ~90 major Russian cities in English; type any custom city to add it.
💯 Number of places to extract	Max items per (query × city) pair.
🟢 Language	Yandex regional domain: RU → yandex.ru, BY → yandex.by, EN → yandex.com, KK → yandex.kz, TR → yandex.com.tr, AZ → yandex.az. Proxy country auto-paired.

🔧 Filters & categories (optional)

Field	Effect
Rating	Minimum average rating threshold: Any · 3.0+ · 3.5+ · 4.0+ · 4.5+ · 4.7+ · 4.9+.
Price categories	Multi-select: Low · Average · Above average · High.
Good place	Only places with rating ≥ 4.5 and ≥ 50 reviews.
Categories	100+ presets (Carwash, Pizzeria, Dental clinic, Driving school…). Bilingual matching — works with Yandex's Russian category names automatically.
Append to global dataset	Also append unique places to a persistent global dataset (cross-run dedupe).

🖼 Photos (optional, paid)

Field	Effect
Number of additional images per place	Extra photo URLs collected. 0 disables. The main image is always free.
Include image authors	Add author metadata for each extra photo.

⭐ Reviews (optional, paid)

Field	Effect
Number of reviews per place	Reviews fetched from each place's /reviews/ page. 0 disables.
Reviews sorting	By default · Newest first · Highest rating first · Lowest rating first.
Include reviewers' data	Author profile fields (name, avatar, public ID) on each review.
📧 Email & phones from website	Visits each place's website, harvests mailto: / tel: links plus inline emails from /contacts and /contact pages → fills emails and websitePhones fields.

🍽 Collect menu (optional)

Field	Effect
Collect menu	Pull menu sections and items where Yandex exposes them (mostly restaurants).

🏢 Scrape with place IDs (optional)

Field	Effect
Organization IDs	Skip search entirely — open these Yandex business OIDs directly. Reviews / photos / contacts still apply per place.

---

Output

Each scraped place becomes one dataset record with 60+ fields. The high-value ones:

Field	Example
title / shortTitle	"PizzaFabrika"
categoryName / categories	"Pizzeria" · ["Pizzeria", "Cafe"]
address / fullAddress	"Moscow, Mikhaylova Street, 30Ak2"
phone / phoneUnformatted / phones	"+7 800 550-06-00"
website / urls	"https://pizzafabrika.ru"
emails (opt-in)	["info@pizzafabrika.ru"]
websitePhones (opt-in)	["+78005500600"]
totalScore / ratingCount / reviewsCount	4.9 · 350 · 116
reviews (opt-in)	[{author, rating, text, updatedTime, businessReply, ...}]
location	{ "lat": 55.84, "lng": 37.62 }
mainPhotoUrl	"https://avatars.mds.yandex.net/.../XXL"
photos / photoDetails (opt-in)	extra photo URLs + alt/id metadata
workingHours / openingHoursText / isOpenNow	day-by-day schedule + status
featureDetails	structured features (Wi-Fi, parking, kids-friendly…)
metroDetails / stopDetails	nearby metro / transit stops with distances
socialLinks	Instagram / VK / Telegram links
menu / menuCategories / menuItemCount	menu data (when Yandex exposes it)
placeId / yandexUri / url	Yandex business ID, internal URI, canonical URL
chain / awards / parentRubrics	brand info, awards, broader rubrics
panorama / panoramaPreviewUrl	nearby panorama
bounds / entrances	geometry corners and known entrances
scrapedAt	ISO timestamp of extraction

The Output tab in Apify Console gives you pre-built table views: Overview · Contacts · Ratings · Features · Media · Menu · Reviews · All fields.

---

Tips & gotchas

• Russian cities work in either spelling. Type Moscow or Москва — Yandex resolves both.
• Belarusian language — pick BY when scraping Minsk / Gomel / Brest / Grodno to route through yandex.by + Belarusian residential IPs. Drastically lowers captcha rate vs running on yandex.ru.
• Rating 4.9+ is strict. Most popular places sit at 4.5–4.8. If you get 0 results, lower the threshold.
• maxItems × locations × queries is your upper bound on records. Need exactly 100? Use one query, one location, maxItems: 100.
• Reviews & photos are opt-in extras. Leave them at 0 if you only need names / addresses / phones / emails.
• Captchas are rare thanks to TLS impersonation + per-language residential proxies. When they do happen, the actor auto-rotates and retries up to 5 times.

---

FAQ

Do you support cities outside Russia? Yes. The dropdown is curated for Russian cities, but you can type any city — Minsk, Almaty, Istanbul, Ankara, Baku — and it works. Pick the matching Language for best results.

Is the proxy configurable? Not needed — proxy country is auto-paired with Language (RU → RU, BY → BY, KK → KZ, TR → TR, AZ → AZ, EN → US). This setup minimizes captcha rate. If you have a special use case, contact me.

What if a place has no website? The emails and websitePhones fields stay null. You still pay only for the place itself ($3 / 1,000), not for the attempted contact scrape.

What if a place has zero reviews? The actor sees reviewsCount=0 from Yandex and skips the reviews fetch entirely — you don't pay the review billing event for empty results.

Can I limit total spend per run? Yes — set maxTotalChargeUsd in Run Options. The actor honors it and stops before exceeding.

How fast is it? ~150 places per minute on a single run with reviews + contacts off. Add 0.5–1.5s per place when contacts are on (depends on the place's website speed).

My output is empty — why? Check the Log tab. Common causes:

• Empty Search queries + empty Categories + empty Organization IDs → nothing to do.
• Too-strict filters (Rating 4.9+ on a small city, Good place + niche category).
• Russian-language city while Language=TR → yandex.com.tr doesn't index Russian cities. Use RU instead.

How do I run this from the API? Click the API button in Console — copy ready-to-paste cURL or SDK snippets.

What about reviews for non-Russian places? Pass the Language matching the region (KK for Kazakh, TR for Turkish, etc.). Reviews are scraped from each place's /reviews/ page on the same Yandex regional domain.

---

Author

Built and maintained by kitcunemia.

Open for custom scraping projects and bulk one-off extractions. Need a different source covered (Google Maps, 2GIS, OpenStreetMap)? Send a message via my Apify profile — typical turnaround is a few days.

---

Changelog

• v0.7 — Belarusian language support, parallel reviews+contacts, reviewsCount=0 short-circuit, slimmer contacts paths, faster timeouts.
• v0.6 — Pay-per-event billing, email/phone scraping, bilingual categories, English UI.
• v0.5 — Initial release: 60+ fields, Apify-proxy auto-rotation, TLS impersonation.