Sreality.cz Real Estate Scraper

Extract property listings from sreality.cz — Czech Republic's #1 real estate portal (Seznam.cz). Scrape apartments, houses, land, commercial space and other property categories by region, transaction type (sale / rent / auction), subtype, price (CZK), surface area and land area.

---

How It Works

Sreality exposes a public, undocumented JSON API at /api/cs/v2/estates. The actor fetches it directly via got-scraping, paginates by page + per_page, and maps the embedded estates[] array into flat records. No HTML, no browser, no cheerio — purely structured JSON in/out.

GET https://www.sreality.cz/api/cs/v2/estates ?category_main_cb={1=apt|2=house|3=land|4=commercial|5=other} &category_type_cb={1=sale|2=rent|3=auction} &category_sub_cb={pipe|sep|sub|codes} &locality_country_id=10001 &locality_region_id={1..14} &czk_price_from=&czk_price_to= &usable_area_from=&usable_area_to= &estate_area_from=&estate_area_to= &per_page=60&page={N}&sort={0|1|2}

Sreality's API does not deploy heavy bot defense (no DataDome / Reese84 / Akamai). Datacenter Apify Proxy works at low volume; RESIDENTIAL with country=CZ is recommended for sustained scraping.

---

Input

{
  "regionIds": ["10"],
  "transaction": "sale",
  "category": "apartment",
  "subtypeCodes": ["4", "5", "6", "7"],
  "priceMin": 3000000,
  "priceMax": 8000000,
  "usableAreaMin": 40,
  "usableAreaMax": 120,
  "estateAreaMin": 0,
  "estateAreaMax": 0,
  "sort": "newest",
  "perPage": 60,
  "maxListings": 500,
  "maxPagesPerTask": 20,
  "requestDelay": 600,
  "maxRetries": 3,
  "proxyConfiguration": { "useApifyProxy": true, "apifyProxyGroups": ["RESIDENTIAL"], "apifyProxyCountry": "CZ" }
}

Parameters

Parameter	Type	Default	Description
regionIds	string[]	required	locality_region_id values. Each becomes its own task.
transaction	enum	sale	sale / rent / auction (maps to category_type_cb)
category	enum	apartment	apartment / house / land / commercial / other (maps to category_main_cb)
subtypeCodes	string[]	[]	Optional category_sub_cb codes (pipe-joined in request)
priceMin / priceMax	int	0	CZK range (0 = no bound)
usableAreaMin / usableAreaMax	int	0	Usable m² range
estateAreaMin / estateAreaMax	int	0	Land/plot m² range
sort	enum	newest	newest / price_asc / price_desc
perPage	int	60	Page size (max 100, 60 safest)
maxListings	int	200	Total cap across all region tasks (0 = unlimited)
maxPagesPerTask	int	10	Pages per region. Sreality loops at ~page 100 on broad filters.
requestDelay	int (ms)	600	Delay between page fetches
maxRetries	int	3	Retries on HTTP errors (rotates proxy IP)
proxyConfiguration	object	RESIDENTIAL+CZ	RESIDENTIAL recommended; datacenter usually OK at low volume

---

Region IDs (locality_region_id)

Region	ID
Hlavní město Praha	10
Středočeský kraj	11
Jihočeský kraj	13
Plzeňský kraj	5
Karlovarský kraj	2
Ústecký kraj	9
Liberecký kraj	4
Královéhradecký kraj	3
Pardubický kraj	6
Kraj Vysočina	14
Jihomoravský kraj	1
Olomoucký kraj	12
Zlínský kraj	15
Moravskoslezský kraj	7

The above IDs are best-effort based on community references. To verify any ID, open sreality.cz, pick the region in the filter, and inspect the network tab — the request URL contains locality_region_id={ID}.

---

Subtype Codes (category_sub_cb)

Apartments (category=apartment, category_main_cb=1)

Code	Subtype
2	1+kk
3	1+1
4	2+kk
5	2+1
6	3+kk
7	3+1
8	4+kk
9	4+1
10	5+kk
11	5+1
12	6 a více
16	atypický

Houses (category=house, category_main_cb=2)

Code	Subtype
33	rodinný (family)
35	vila
37	chalupa (cottage)
39	chata (chalet)
40	na klíč
43	usedlost (homestead)
47	zemědělská usedlost
48	památka (heritage)

Land (category=land, category_main_cb=3)

Code	Subtype
18	bydlení (residential)
19	komerční
20	pole (field)
21	louka (meadow)
22	les (forest)
23	rybník (pond)
24	sad (orchard)
25	zahrada (garden)
26	ostatní

Commercial (category=commercial, category_main_cb=4)

Code	Subtype
27	kanceláře (offices)
28	sklady (warehouses)
29	výroba (production)
30	obchodní prostory (retail)
31	ubytování (accommodation)
32	restaurace
41	zemědělský objekt
50	činžovní dům
51	ostatní

Other (category=other, category_main_cb=5)

Code	Subtype
13	garáž
14	vinný sklep
15	půdní prostor
17	ostatní

---

Output

Example record:

{
  "adId": "2614321500",
  "detailUrl": "https://www.sreality.cz/detail/prodej/byt/2%2Bkk/praha-5-smichov-stroupeznickeho/2614321500",
  "apiUrl": "https://www.sreality.cz/api/cs/v2/estates/2614321500",
  "title": "Prodej bytu 2+kk 50 m²",
  "category": "apartment",
  "categoryCode": 1,
  "transactionType": "sale",
  "transactionCode": 1,
  "subtype": "2+kk",
  "subtypeCode": 4,
  "price": 5500000,
  "priceText": "5 500 000 Kč",
  "priceUnit": "Kč",
  "priceNote": null,
  "currency": "CZK",
  "pricePerSqm": 110000,
  "usableArea": 50,
  "estateArea": null,
  "locality": "Praha 5 - Smíchov, Stroupežnického",
  "region": null,
  "district": "Praha 5 - Smíchov",
  "city": "Praha",
  "latitude": 50.0729,
  "longitude": 14.4083,
  "mainImageUrl": "https://d18-a.sdn.cz/d_18/c_img.../image.jpg",
  "imageUrls": ["https://d18-a.sdn.cz/.../1.jpg", "https://d18-a.sdn.cz/.../2.jpg"],
  "imageCount": 12,
  "hasFloorPlan": true,
  "hasVideo": false,
  "has3dTour": false,
  "agentId": "1234",
  "agentName": "Reality XYZ s.r.o.",
  "agentLogoUrl": "https://d18-a.sdn.cz/.../logo.png",
  "agentUrl": "https://www.sreality.cz/firma/reality-xyz/1234",
  "isExclusive": true,
  "isNew": false,
  "isAuction": false,
  "isTopped": false,
  "labels": ["EXCLUSIVE_OFFER"],
  "seoLocality": "smichov-praha-5",
  "searchRegionId": 10,
  "searchCategory": "apartment",
  "searchTransaction": "sale",
  "searchSubtypes": [4],
  "scrapedAt": "2026-05-04T13:00:00.000Z"
}

Notes

• adId is sreality's stable hash_id (numeric, but stored as string).
• detailUrl is built from seo.{category_type_cb,category_main_cb,category_sub_cb,locality} + hash_id. If any slug is missing the URL will still resolve — sreality redirects on hash_id alone.
• apiUrl is the JSON detail endpoint — useful for v1.1 enrichment (fetchDetails flag, planned).
• pricePerSqm is computed (price / usableArea); not provided by the API directly.
• region / district / city are best-effort parsed from the locality string. For full administrative hierarchy use the detail endpoint (planned).
• labels contain raw API labels like EXCLUSIVE_OFFER, NEW, LOWERED_PRICE, RECOMMENDED.

---

Use Cases

• Market research — CZK/m² heatmaps by region, district and subtype
• Investment screening — yield = monthly rent / sale price per area
• Lead generation — discover all active agencies via agentId cluster
• Price tracking — re-run periodically and diff against prior datasets
• Auction monitoring — set transaction: auction to track foreclosure inventory
• Land scouting — category: land + estateAreaMin/Max for plot search
• New-listings feed — sort: newest + maxPagesPerTask: 1 for daily delta

---

Cost & Performance

• Architecture: pure HTTP + JSON parse. No browser, no DOM, no images downloaded.
• Throughput: 60 listings per page, ≈ 1-2 pages/sec at default requestDelay=600. ≈ 200 listings in 4-8s.
• Memory: 256MB minimum.
• Compute Units: ~0.002 CU per 100 listings.

---

Limitations

• Listing-level data only in v1.0. Detail-only fields (full description, energy class, floor number, year built, full amenities list, all images at full resolution) require a second per-item API call — planned for v1.1 behind a fetchDetails flag.
• Pagination ceiling — sreality's API loops at ~100 pages on broad filters (returns the same first page repeatedly). The actor detects this loop (compares the first hash_id across pages) and stops. Narrow with priceMin/Max or usableAreaMin/Max for full coverage of saturated regions.
• Subtype codes are documented best-effort. If a code returns 0 results you may need to verify it via DevTools on a live filter.
• Locality parsing (city / district) uses heuristics on the free-text locality field; for guaranteed accuracy use the structured locality IDs from the detail endpoint (planned for v1.1).

---

Changelog

Version	Date	Notes
1.0.0	2026-05-04	Initial release — direct JSON API scraping via /api/cs/v2/estates, 14 regions, 5 categories, 3 transaction types, 40+ subtype codes, full pagination with loop detection, image gallery extraction, agency embedding, structured price/area/GPS output.