Yandex Maps Scraper

Search public Yandex Maps business listings by keyword and location. Version 3 reads Yandex's server-rendered structured result state over HTTP, validates the response before parsing, and uses a narrowly scoped HTML link fallback if that structure changes. It does not store transport IDs, tokens, cookies, challenge pages, or raw responses.

Modes

• Single search: set query.
• Batch search: set queries; records are selected round-robin so one query cannot consume the global limit.
• Single item: set maxItems to 1.
• Pagination: raise maxItems and maxPages (up to 10 pages per query).
• Filters: category text, rating, review count, open-now, phone, website, and photo availability.
• Sorting: Yandex relevance, rating, review count, or name.

location is combined with the query exactly as a person would enter it in Maps. searchDomain and language control the Yandex surface. country is a provenance label; use proxyCountryCode when network geolocation matters.

Proxy support

The Actor supports direct HTTP, Apify Proxy, country selection, proxy groups, and custom proxy URLs. GOOGLE_SERP is rejected because Apify's SERP proxy is for supported Google Search/Shopping HTTP requests, not Yandex Maps. Residential proxies may improve reliability, but the Actor does not solve or bypass CAPTCHAs or other access controls.

Output

The stable dataset includes place IDs and canonical Maps URLs; names and addresses; coordinates; categories; public phones, websites, and social/business links; rating/review values; current and weekly hours; photos; public feature metadata; region data; query/page provenance; and extraction timestamps. Optional values absent from Yandex are omitted rather than stored as null, empty strings, or empty arrays.

Local validation

npm ci
npm test
node validate-datasets.js
apify run --purge --input-file qa-inputs/local-single.json

The representative single, batch, pagination, filtered, empty-result, invalid-input, low-concurrency, proxy-disabled, and Residential-cloud inputs are under qa-inputs/.

Limitations

Results vary with Yandex ranking, language, domain, location wording, network region, and public business metadata. Filters are applied only to source-provided values. If Yandex returns a challenge or an unrecognized page, the Actor fails without publishing a partial or fabricated dataset.