Property AU Scraper

Scrape Australian property listings from property.com.au with rich detail-page enrichment. Each listing record comes back with far more than the usual "address + price + beds". You get the full picture a buyer would see on the site.

What you get per listing

Every result is a complete property record:

• Address & location: full street address, suburb, state, postcode, latitude & longitude
• Listing state: for sale / for rent / sold / off-market, plus the current asking price and price detail (e.g. "Listed by agent on a given date")
• Core features: bedrooms, bathrooms, car spaces, land size, floor area, year built, mobile coverage
• Full description: the listing's long-form description text
• Price history timeline: every sale, rent, lease and listing event on the property, often going back decades
• AVM value estimates: automated sale and rental valuations with confidence bands
• Nearby schools: zoned and non-zoned primary & secondary schools with distance, student counts, sector and grades
• Planning overlays: bushfire, flood and heritage overlay flags, council name, land description
• NBN / internet availability: connection type (FTTP / FTTB / FTTN / HFC), quality rating, description
• Market insights: suburb-level median price & rent with 12 month change
• Media: image URLs (all photos), media counts (floor plans, videos, 3D tours)
• Breadcrumbs: the navigation hierarchy from state to suburb to street

Two ways to run it

Listing mode (default). Describe what you want with structured filters:

{
  "mode": "listing",
  "search": "southbank",
  "listing_type": "buy",
  "propertyTypes": ["HOUSE", "APARTMENT"],
  "priceMin": 500000,
  "priceMax": 900000,
  "bedroomsMin": 2,
  "max_properties": 100,
  "max_pages": 5
}

search accepts a suburb name, postcode, or city: southbank, 3006, brisbane, melbourne. Ambiguous or misspelled queries fail fast rather than running against the wrong location.

URL mode. Paste property.com.au search URLs directly and let the scraper walk them forward:

{
  "mode": "url",
  "urls": [
    "https://www.property.com.au/search/?locations=Melbourne%2C+VIC+3000&propertyStatus=FOR_SALE&pageNumber=2"
  ],
  "max_properties": 60,
  "max_pages": 3
}

Pagination walks forward from the URL's own ?pageNumber=N. Pass pageNumber=2 with max_pages=3 and you get pages 2, 3 and 4.

Filters (listing mode)

Field	Notes
listing_type	buy, rent, or sold
propertyTypes	HOUSE, APARTMENT, TOWNHOUSE, VILLA, LAND, ACREAGE, RURAL, BLOCK_OF_UNITS, RETIREMENT
priceMin / priceMax	AUD
bedroomsMin / bathroomsMin
landSizeMin	square metres
carSpacesMin
keywords	e.g. "pool", "renovated", "corner block"
excludeUnderOffer	buy only
sortBy	NEW_DESC (default), PRICE_ASC, PRICE_DESC, RELEVANCY, NEXT_AUCTION, NEXT_INSPECTION

In URL mode these filters are ignored. Encode them directly in your URLs (e.g. &minPrice=500000&minBedrooms=2).

Limits & dataset

• max_properties: hard cap on results (0 = unlimited)
• max_pages: per URL page budget (0 = unlimited)
• dataset_name: custom dataset name; omit or use default for the default dataset
• clear_dataset: wipe the dataset before the run

Each listing is pushed into the run's dataset as a JSON record. A run summary is written to the key value store under METADATA with mode, total scraped, duration and throughput.

Example output record (shape only)

{
  "id": "<property id>",
  "address": "<full street address>",
  "suburb": "<SUBURB>",
  "state": "<STATE>",
  "postcode": "<postcode>",
  "latitude": 0,
  "longitude": 0,
  "listing_type": "buy",
  "listing_status": "<current status>",
  "property_type": "<house | unit | townhouse | ...>",
  "price": "<asking price text>",
  "price_details": "<listing detail text>",
  "features": {
    "bedrooms": 0,
    "bathrooms": 0,
    "car_spaces": 0,
    "land_size": "<area with unit>",
    "floor_area": "<area with unit>",
    "built_year": 0,
    "mobile_coverage": "<coverage tier>"
  },
  "description": "<long form listing description>",
  "timeline": [
    { "event": "listed_for_sale", "year": "NOW", "price": "<price>", "date": "<date>", "agent": "<agent>" },
    { "event": "sold", "year": 0, "price": "<price>", "date": "<date>" },
    { "event": "leased", "year": 0, "date": "<date>" }
  ],
  "value_estimates": {
    "sale": { "value": 0, "display": "<price>", "confidence": "HIGH", "range_min": 0, "range_max": 0 },
    "rental": { "value": 0, "display": "<price pw>", "confidence": "MEDIUM", "range_min": 0, "range_max": 0 }
  },
  "schools": {
    "primary": [
      { "name": "<school name>", "is_zoned": true, "grades": "Prep-6", "sector": "<Government | Catholic | Independent>", "students": 0, "distance": "<km>" }
    ],
    "secondary": [
      { "name": "<school name>", "is_zoned": true, "grades": "7-12", "sector": "<Government | Catholic | Independent>", "students": 0, "distance": "<km>" }
    ]
  },
  "planning_overlays": { "bushfire": false, "flood": false, "heritage": false },
  "council": "<council name>",
  "land_description": "<planning / zoning description>",
  "internet": { "connection_type": "<FTTP | FTTB | FTTN | HFC>", "quality": "<quality tier>" },
  "market_insights": {
    "median_price": "<suburb median>",
    "price_growth_12mo": "<percent>",
    "median_rent": "<suburb median rent>",
    "rent_growth_12mo": "<percent>"
  },
  "media": { "image_count": 0, "images": ["<image url>"] },
  "breadcrumbs": [{ "title": "<segment>", "url": "<link>" }],
  "property_url": "<canonical property page url>",
  "status": "success"
}

Tips

• Start small. Set max_properties: 10 for a first run to verify your filters before scaling up.
• Use URL mode for complex filters. Property.com.au's own search URL encodes every filter the site supports. Build a search on the site, copy the URL, and paste it into urls.
• max_pages is a per URL budget. If you pass three URLs and set max_pages: 5, the scraper walks up to 5 pages for each of them.
• Sold listings show "Off market" status. That's how property.com.au displays them. The sold price sits inside the timeline array.