Hemnet Property Scraper

Bulk-scrape property listings from hemnet.se, Sweden's largest housing marketplace: homes for sale, completed sales with final prices (slutpriser), and upcoming listings (kommande) announced before they go live. Feed it a Hemnet search URL, a link to a single property, or a list of Swedish place names with a filter template. You get complete listing data, down to broker contacts, amenities and floor plans.

Search URLs and links to individual properties both work, in the same field. Paste hemnet.se/bostader?… for a search, or hemnet.se/bostad/…-21730633 for one specific home. Each link becomes one row, with no search to page through.

No headless browser. It runs on direct HTTP requests to Hemnet's own mobile API, so runs finish fast and cost a fraction of a browser-based scraper.

Common use cases

• Market monitoring: daily delta of new listings in selected municipalities using the onlyNew flag
• Sold-price comps: completed sales with final price, and the gap from the asking price in both kronor and percent
• Investment research: price per m², monthly fee, running costs, construction year and energy class across several areas in one run
• Broker lead generation: name, email, phone and agency for every listing, filtered by area and property type
• New-build tracking: the Kommande section plus isDevelopment and isNewConstruction to catch projects before they reach the open market
• Distressed assets: Hemnet's own foreclosure filter is read from the for-sale URL, and isForeclosure marks Kronofogden (enforcement authority) sales in the output
• Aggregator feeds: bulk extraction for portals combining Hemnet with other sources

How to use it

Three input formats are supported, and they can be combined in one run.

Option 1: paste search URLs from Hemnet

Build the search on hemnet.se with whatever filters you like: price, rooms, living area, property type, balcony, elevator, keywords, ground floor, expanded area. Then paste the address bar into urls. The filters and the section are both read from the URL: /bostader for sale, /salda/bostader for sold prices, /kommande/bostader for upcoming.

{
  "urls": [
    "https://www.hemnet.se/bostader?location_ids%5B%5D=17744&item_types%5B%5D=bostadsratt&price_max=5000000",
    "https://www.hemnet.se/salda/bostader?location_ids%5B%5D=18031&sold_age=6m"
  ],
  "searchLimit": 50,
  "onlyNew": false,
  "proxySettings": { "useApifyProxy": false }
}

Some filters can't be applied. Hemnet's sold-price search supports fewer of them than the for-sale one, for instance. The run log then says exactly which filter was dropped. Nothing is ignored silently.

Option 2: paste links to individual properties

Drop them into the same urls field, mixed with searches if you like. A home for sale, an upcoming one and a completed sale all work, and each is fetched on its own with no search around it.

{
  "urls": [
    "https://www.hemnet.se/bostad/lagenhet-2rum-centrum-halmstads-kommun-vattugatan-2a-21730633",
    "https://www.hemnet.se/salda/lagenhet-2rum-norrmalm-skovde-kommun-kansligatan-7b-1311609520551872240"
  ],
  "proxySettings": { "useApifyProxy": false }
}

Option 3: apply one filter template across many locations

When you need the same filters across many places, pass one example URL as sampleUrl and the location list separately. Its filters and its section are applied to every location; its own location is ignored. Without a sampleUrl, locations are searched for sale.

{
  "locations": ["Stockholms län", "Göteborg", "Nacka kommun"],
  "sampleUrl": "https://www.hemnet.se/bostader?rooms_min=3&price_max=6000000&item_types%5B%5D=bostadsratt",
  "searchLimit": 100,
  "onlyNew": false,
  "proxySettings": { "useApifyProxy": false }
}

Locations are checked before anything is scraped, and named in the log (17744 is Stockholms län (COUNTY)). This matters because Hemnet itself ignores a location it doesn't recognise and answers with the whole country instead. So an old URL whose area has since been redrawn would quietly return all of Sweden. The actor refuses that search rather than billing you for it.

Other parameters

• searchLimit: maximum results per search, applied to each URL and each location separately. Set to 0 for all results. Ignored when onlyNew is enabled.
• onlyNew: returns only what has appeared since the previous run. New listings, or newly registered sales when the search is a /salda section. The first run with this flag does a full scan and records its position. Later runs stop as soon as they reach it. Works for all three sections, and the filters in your URL still apply, so a sold-date window is a good way to keep that first scan small. searchLimit is the one exception: it's ignored here, because a cap would move the position past listings it never returned.
• stateStore: key-value store holding the onlyNew position, one per search. Leave empty to use the auto-created hemnet-tech-store. Pick a separate store per monitoring task to keep their positions isolated. To reset monitoring and force a full scan, point it at a different store.
• proxySettings: standard Apify proxy configuration.

What Only New Listings guarantees

Within a run you're never charged twice for the same row.

Across runs a repeat is possible. A run cut short keeps its previous position, so the next one may re-deliver a few rows you already have. Nothing is skipped. Match on id if your pipeline needs to be certain.

Sales registered late by the broker can be missed. Run Sold without onlyNew when you need every sale.

Hemnet's 2 500-result window

Hemnet serves at most 2 500 results per search. It's the site's own limit, not the actor's.

Split a wide search into narrower ones and pass several URLs: by municipality, by price band, or by property type. The log says how many results a search left behind, so you know which one to split.

Output

Data is saved to an Apify dataset with seven preconfigured views: Overview, For Sale, Upcoming, Sold Prices, Location Data, Brokers and Full Data. Each view picks the columns that section fills. Overview, For Sale, Upcoming and Sold Prices lead with a photo, so the table reads as a gallery rather than a grid of ids. Download in JSON, CSV, XML, Excel, HTML, RSS or JSONL.

Every value arrives ready to use rather than ready to display. Prices are whole kronor as integers, not "2 195 000 kr". Areas and room counts are numbers, not "70 m²" and "2,5 rum". Timestamps are ISO-8601 UTC. The dataset drops into a spreadsheet, a database or a model without a parsing step.

Sample output: property for sale

{
  "id": "21730633",
  "searchMode": "for_sale",
  "url": "https://www.hemnet.se/bostad/lagenhet-2rum-centrum-halmstads-kommun-vattugatan-2a-21730633",
  "streetAddress": "Vattugatan 2A",
  "area": "Centrum",
  "districts": ["Mickedala", "Halmstad"],
  "municipality": "Halmstads kommun",
  "county": "Hallands län",
  "latitude": 56.6801949,
  "longitude": 12.8557949,
  "housingForm": "Lägenhet",
  "tenure": "Bostadsrätt",
  "askingPriceSek": 2195000,
  "priceOnRequest": false,
  "feeSek": 4272,
  "runningCostsSek": 3314,
  "squareMeterPriceSek": 31357,
  "numberOfRooms": 2,
  "livingAreaSqm": 70,
  "floorNumber": 2,
  "floorsInBuilding": 5,
  "constructionYear": 1989,
  "energyClass": "D",
  "publishedAt": "2026-08-02T13:17:11.123000+00:00",
  "daysOnHemnet": 0,
  "timesViewed": 0,
  "biddingStarted": false,
  "closestWaterDistanceMeters": 350,
  "coastlineDistanceMeters": 2100,
  "labels": ["Tis 11 aug kl 17:00", "Premium", "Balkong", "Hiss"],
  "adTier": "PREMIUM",
  "hasVideo": false,
  "has3dTour": false,
  "amenities": { "Balkong": true, "Hiss": true },
  "openHouses": [
    {
      "start": "2026-08-11T15:00:00+00:00",
      "end": "2026-08-11T15:30:00+00:00",
      "description": "Tidsbokad visning. Boka själv via hemsidan!",
      "isOnlyDate": false
    }
  ],
  "housingCooperative": "BRF Poppeln 16",
  "housingCooperativeRegistrationNumber": "7164082468",
  "housingCooperativeRegisteredYear": 1986,
  "mainImageUrl": "https://bilder.hemnet.se/images/itemgallery_L/6a/e4/6ae43aa3….jpg",
  "imageCount": 32,
  "imageUrls": ["https://bilder.hemnet.se/images/itemgallery_L/…", "…31 more"],
  "floorPlanImageUrls": ["https://bilder.hemnet.se/images/itemgallery_L/…"],
  "brokerName": "Claudia R…",
  "brokerEmail": "claudia.ryden@…",
  "brokerPhone": "0735-18****",
  "brokerAgencyName": "MOHV Halmstad",
  "brokerAgencyCity": "Halmstad",
  "brokerAgencyUrl": "http://www.mohv.se/"
}

Sample output: completed sale

{
  "id": "1311609520551872240",
  "searchMode": "sold",
  "url": "https://www.hemnet.se/salda/lagenhet-2rum-norrmalm-skovde-kommun-kansligatan-7b-1311609520551872240",
  "listingId": "21755181",
  "streetAddress": "Kansligatan 7B",
  "districts": ["Norrmalm", "Skövde"],
  "municipality": "Skövde kommun",
  "county": "Västra Götalands län",
  "latitude": 58.39475,
  "longitude": 13.85011,
  "housingForm": "Lägenhet",
  "tenure": "Bostadsrätt",
  "askingPriceSek": 1250000,
  "finalPriceSek": 1200000,
  "priceChangeSek": -50000,
  "priceChangePercent": -4.0,
  "squareMeterPriceSek": 25532,
  "feeSek": 3218,
  "numberOfRooms": 2,
  "livingAreaSqm": 47,
  "constructionYear": 1943,
  "soldAt": "2026-08-02T10:00:00+00:00",
  "labels": ["Balkong"],
  "amenities": { "Balkong": true, "Hiss": false },
  "housingCooperative": "BRF Kansligården",
  "housingCooperativeRegistrationNumber": "7666000224",
  "housingCooperativeRegisteredYear": 1943,
  "mainImageUrl": "https://bilder.hemnet.se/images/itemgallery_L/1c/7b/1c7b7040….jpg",
  "imageCount": 21,
  "imageUrls": ["https://bilder.hemnet.se/images/itemgallery_L/…", "…20 more"]
}

Sample output: upcoming listing

{
  "id": "21760410",
  "searchMode": "upcoming",
  "url": "https://www.hemnet.se/bostad/lagenhet-2rum-bromma-blackeberg-stockholms-kommun-holbergsgatan-90-21760410",
  "streetAddress": "Holbergsgatan 90",
  "area": "Bromma - Blackeberg",
  "districts": ["Blackeberg", "Bromma", "Bromma, västra", "Stockholm"],
  "municipality": "Stockholms kommun",
  "county": "Stockholms län",
  "latitude": 59.34639,
  "longitude": 17.8809948,
  "housingForm": "Lägenhet",
  "tenure": "Bostadsrätt",
  "askingPriceSek": 2295000,
  "feeSek": 4158,
  "runningCostsSek": 6000,
  "squareMeterPriceSek": 46837,
  "numberOfRooms": 2,
  "livingAreaSqm": 49,
  "floorNumber": 6,
  "floorsInBuilding": 9,
  "constructionYear": 2006,
  "energyClass": "E",
  "isUpcoming": true,
  "markedAsNew": true,
  "description": "Välkommen till denna smakfulla och välplanerade tvårummare på våning 6…"
}

Every row carries searchMode: "for_sale", "upcoming" or "sold". Group on it to split a mixed run back into its sections.

The full field list, with a description and a unit for every field, is in the dataset schema on the Storage tab after each run.

Is it legal to scrape Hemnet?

This scraper accesses only data that is publicly visible on hemnet.se without authentication. It doesn't sign in. It doesn't bypass paywalls. It doesn't try to reach anything Hemnet keeps behind one.

Scraping publicly available data is generally legal. See Apify's overview: Is web scraping legal? Yes, if you know the rules. You're responsible for your own use complying with Hemnet's Terms of Service and applicable law.

Three fields are personal data under the GDPR, even though Hemnet publishes them: brokerName, brokerEmail and brokerPhone. Storing or processing them needs a lawful basis. If you don't need them, download any view except Brokers and Full Data, or drop the columns with the omit parameter.

Agency-level contacts are organisational data, not personal data: brokerAgencyName, brokerAgencyCity, brokerAgencyPhone, brokerAgencyEmail and brokerAgencyUrl describe a business, not an individual.

Bugs and feedback

If something doesn't work or a field is missing, the Issues tab is the fastest way to reach me.