Israeli Supermarket Prices - Cross-Chain Compare

Compare any product's price across every major Israeli supermarket chain in one run. Built directly on the price files chains are required to publish by law (the Food Price Transparency regulations), so the data is official, complete, and refreshed daily.

Supported chains (25): Shufersal, Rami Levy, Carrefour, Tiv Taam, Osher Ad, Yochananof, Keshet Taamim, Stop Market, Super Yuda, Fresh Market (Paz), Politzer, Salach Dabach, Hazi Hinam, Shuk Ha'ir, Bareket, King Store, Maayan 2000, Zol VeBegadol, Shefa Birkat Hashem, Super Sapir, Good Pharm, H. Cohen, Wolt Market, Victory, Machsanei HaShuk.

Modes

Mode	What you get	Inputs
compareProducts (default)	The same product's price at every chain, ready to sort cheapest-first	barcodes and/or productQuery, chains, oneStorePerChain
storeSnapshot	Every product in one store	chain, storeId (optional), category (optional)
listStores	Every branch of a chain (id, name, address, city)	chain (or chains)
chainDump	An entire chain (large)	chain, confirmFullDump: true
promoSnapshot	Current promotions, parsed: per-item derived unit price, min quantity, dates, active/club/coupon flags, and a promoPriceTrusted flag (true only when the derived price matches the chain's own promo wording - reward-type promos like "second unit 50% off" publish misleading prices)	chain, storeId (optional)

Comparing a product (the main use)

Match by barcode for an exact, reliable cross-chain comparison (each chain names products differently, so name matching is fuzzier):

{ "mode": "compareProducts", "barcodes": ["7290000042015", "7290010945306"], "oneStorePerChain": true }

Returns one row per chain that carries the barcode, e.g.:

{ "barcode": "7290000042015", "itemName": "חלב תנובה 3% שומן 1 ליטר", "price": 6.20, "chainName": "Rami Levy (רמי לוי)", "storeId": "044", "priceUpdateDate": "2026-06-10T00:10:07" }

Sort the rows by price to see who is cheapest. productQuery (a Hebrew name substring like "חלב 3%") is also supported for discovery.

Output

One row per matched product per chain, with the shelf price and unitPrice (both ILS), barcode, itemName, manufacturer, weight/unit info, the priceUpdateDate, and chainName / chainId / storeId / sourceFile.

How many rows will I get? (cost guide)

One row = one product listing at one chain. Typical counts at the default settings:

Query style	Rows	Example
Barcode compare	~22 (one per chain)	7290110325619
Specific product name	60-400	"חלב 3%", "קוקה קולה 1.5"
Broad single word	thousands	"חלב" matches every product containing the word

For broad queries set maxRowsPerChain (e.g. 50) to keep results representative and the cost bounded.

How it works

Each chain publishes gzipped XML price files on a public portal (Shufersal's own site, the shared Cerberus portal used by Rami Levy and others, Carrefour's site). The actor logs in where needed, downloads the relevant store file per chain, stream-parses it (files are 5-15 MB and never loaded whole into memory), normalizes the different portal dialects (including ZIP-wrapped and UTF-16 files) into one schema, and emits only the rows you asked for. Results are pushed continuously, so a long run never loses progress.

Notes & limits

• Prices in a chain's full price file refresh daily (early morning). Intra-day changes land in the incremental files (a future mode).
• oneStorePerChain is on by default because branch-to-branch prices within a chain are nearly identical; turn it off to scan every store.
• chainDump is gated behind confirmFullDump because a whole chain is hundreds of thousands of items.
• Weighed produce uses internal codes, not universal barcodes, so it does not always join across chains (the same internal code can mean different products at different chains - never join weighed items on barcode).
• listStores resolves branch cities to clean Hebrew names (chains publish numeric CBS codes or hide the city in the branch name; the actor handles both).

Use cases

• Cheapest-basket / price-comparison apps and bots.
• Retail and CPG price intelligence (track your product, or a competitor's, across all chains).
• Inflation and consumer-price research.
• Building a normalized national price dataset.