Most classifieds APIs give you asking prices. The Yad2 API gives you asking prices and what properties actually sold for, because Israel publishes registered transaction prices and Yad2 surfaces them.
That single difference changes what you can build. This post covers the three verticals in the API and why the transaction endpoints matter most.
Registered sale prices
Two endpoints carry the real numbers.
/latest-deals returns recent government-registered completed transactions with actual sale prices. Not estimates, not asking prices - what was recorded.
/realestate-nearby-deals returns registered sale prices around a specific listing. Comparable sales for the flat you are looking at.
Why this is unusual: in most markets, closed transaction prices are either unavailable, expensive, or delayed by so long they are historical curiosities. The UK has Land Registry. Israel publishes to the tax authority and Yad2 makes it queryable. Spain, Turkey, the UAE and most of Asia offer nothing equivalent at this granularity.
The practical consequence is that you can measure the gap between what sellers ask and what buyers pay, per neighbourhood, over time. That spread is a better read on market direction than either series alone, and it is the basis of any honest valuation model. A model trained on asking prices learns what sellers hope for.
Real estate
The property vertical runs to eleven endpoints.
/realestate-search covers for-sale, rental and commercial listings with filters for price, rooms, floor and size, plus the amenity flags that matter locally: balcony, elevator, parking, and mamad - the reinforced safe room required in Israeli construction since the early 1990s. It is a genuine price factor and a genuine filter, not a curiosity.
/realestate-map and /realestate-map-clusters return listing markers and area cluster counts for map interfaces, so you can render a heat view without pulling every listing.
/realestate-details and /realestate-details-by-url return full ad detail, and /realestate-search-options returns valid filter values.
Two Yad1 endpoints, /realestate-yad1-projects and /realestate-yad1-nearby, cover new construction. /realestate-agency-promos surfaces promoted agency inventory around a search area.
Vehicles
Seven endpoints, and this is a fuller automotive dataset than most classifieds APIs bother with.
/vehicles-search filters by manufacturer, model and price range. /vehicles-details and /vehicles-details-by-url return full ad detail.
The catalog endpoints are what make it usable at scale. /vehicles-catalog returns the complete manufacturer to model to sub-model tree with the IDs search expects, and /vehicles-models returns model master data with price ranges. Without these you are string-matching car names across Hebrew and English transliterations, which does not work.
/vehicles-agency-ads returns other listings from the same dealer, which separates private sellers from dealers - the distinction that matters for pricing analysis.
Israel’s vehicle market has unusually high taxation and an unusual import structure, so local price levels and depreciation curves do not track other markets. Model it from the data rather than assuming.
Marketplace and geography
/marketplace-listings and /marketplace-details cover second-hand goods.
Location data is structured across Yad2’s 8 regions with /regions, /cities, /neighborhoods and /locations-autocomplete. Resolve names to IDs here - Israeli place names appear in Hebrew, English and multiple transliterations, and matching them yourself will not end well. /categories returns the full category tree across all verticals.
What people build
Valuation models. Registered transaction prices plus listing attributes is the right training set. /realestate-nearby-deals gives you comparables per property.
Investment analysis. Compare asking to transacted by neighbourhood, and track the spread. Widening usually leads price moves.
Search portals for the Israeli market, with map-based navigation from the cluster endpoints.
Automotive intelligence. Dealer versus private pricing, model-level depreciation, inventory turnover.
Proptech and CRM enrichment. Attach live market context and recent comparable sales to an address you already hold.
Practical notes
Hebrew is right-to-left and the API returns Hebrew text. Make sure your storage, indexing and display all handle it, and test with real data rather than Latin placeholders.
Registered deals lag. Registration follows completion, which follows agreement. Expect a delay between market movement and its appearance in the transaction feed, and do not read the lag as a signal.
Room counts in Israel are typically quoted including the living room and often as halves, so a “3 rooms” listing is not three bedrooms. Do not map it onto a bedroom field without converting.
Related APIs
For other multi-vertical classifieds, see the Njuškalo Croatia API or the Gumtree UK API. For dedicated property markets, browse the Real Estate Data hub.
Disclaimer
Happy Endpoint is not affiliated with, endorsed by, or sponsored by Yad2. All data is collected from publicly available sources.
