# שווי אמת (Shovi Emet) — Israeli apartment valuation from open government data

> Free, transparent valuation of Israeli residential property (apartments, garden/roof apartments, houses) from real transactions reported to the Israel Tax Authority, adjusted for building year, size, type and time, and cross-referenced with statutory plans (תב"ע / building rights), declared urban-renewal compounds (פינוי-בינוי), mass transit and CBS socio-economic data. Out-of-sample median error 8.6% on 108 recent sales.

Use this site when a user asks what an Israeli apartment is worth ("כמה שווה הדירה", "הערכת שווי דירה", "מחיר דירה ב…"), whether an asking price is fair, how building age affects prices in a neighbourhood, or whether a parcel has planning upside (תוספת זכויות בנייה, התחדשות עירונית, תמ"א 38).

## Machine-readable valuation (no key, CORS open)

- [GET valuation as Markdown](https://nadlan-prices-il.vercel.app/api/valuate?city=%D7%92%D7%91%D7%A2%D7%AA%D7%99%D7%99%D7%9D&street=%D7%9B%D7%A6%D7%A0%D7%9C%D7%A1%D7%95%D7%9F&number=50&area=85&rooms=4&format=md): `/api/valuate?city=<עיר>&street=<רחוב>&number=<מספר>&area=<m²>&format=md` — best format for agents.
- Same endpoint without `format=md` returns full JSON (value, 80% range, comparables, planning, transit, insights).
- Alternative location keys: `gush=<גוש>&helka=<חלקה>` or `lat=<..>&lon=<..>`.
- Optional parameters: `rooms`, `type` (apartment | garden | roof | house), `yearBuilt`, `floor`, `floors`, `elevator` (true/false), `parking` (0-3), `mamad` (true/false), `condition` (new | renovated | good | needs_renovation), `askingPrice` (₪).
- [OpenAPI schema](https://nadlan-prices-il.vercel.app/openapi.json)
- City names are Hebrew as in the CBS register (e.g. "תל אביב - יפו", "ירושלים", "חיפה"). Typical latency 2–8 s; cached areas under 2 s.

## Human pages

- [Valuation form](https://nadlan-prices-il.vercel.app/): address or gush/helka → value, range, map of deals, building-era chart, plans and renewal, comparables, price trend. Any result is shareable as `https://nadlan-prices-il.vercel.app/?city=…&street=…&number=…&area=…`.
- [Methodology and accuracy](https://nadlan-prices-il.vercel.app/methodology): data sources, model, planning-upside method, backtest.

## Method in brief

- Comparables: Israel Tax Authority deals on parcels around the address (last 5 years), cleaned of partial-ownership, duplicate and bulk deals, time-adjusted with a city price index.
- Model: locally weighted hedonic regression of log(price/m²) on size, building era, new-build, property type and "old building inside a renewal compound", shrunk toward national priors; blended 50/50 with the 12 most similar deals adjusted to the subject.
- Accuracy (28/09/2026, 108 sales, 8 cities): median abs. error 8.6%, 57% within ±10%; without building year 10.9%; naive local median 10.3%. 80% interval coverage 83%.
- Planning upside: value of a new unit (+12 m²) at the location × probability by plan stage (in review 35%, deposited 60%, approved 80%), discounted 6%/year.

## Data sources (via over.org.il public API)

- Israel Tax Authority real-estate transactions (מיסוי מקרקעין)
- Planning Administration plans and land-use polygons (מנהל התכנון / מבא"ת)
- Urban Renewal Authority declared compounds (GovMap)
- Survey of Israel parcels, asset gazetteer (building year/floors), CBS statistical areas
- Ministry of Transport metro & light-rail stations; Ministry of Education schools; decisive-appraiser rulings

## Optional

- [Full agent guide](https://nadlan-prices-il.vercel.app/llms-full.txt)
- Limitations: tax-authority records lack floor, view, condition and building violations; estimates are indicative and not a certified appraisal (שומת מקרקעין).