All company data is sourced from the OCCSZ (Országos Cégnyilvántartó és Céginformációs Szolgálat — National Company Register and Company Information Service), which operates the official Hungarian company register. The system interacts with two distinct services:
OCCSZ XML API — A paid, authenticated XML API providing company search by name, registration number, and tax number. Returns structured XML with company name, registration number, address, tax number, and status. Used for all search operations and as the entry point for company identification.
Trade Register Extract (Cégkivonat) — A comprehensive PDF document purchased from the OCCSZ API for each company. This is the primary source for all structured company profile data. The PDF is parsed deterministically to extract: legal name, legal form, registration date, share capital, tax number, registered address, legal representatives, auditors, supervisory board members, shareholders, activity codes (TEÁOR), establishments, and company status.
Company profile is derived from the Trade Register Extract. The TRE PDF is
purchased from the OCCSZ API, then parsed deterministically. AI enrichment is
applied in parallel for address parsing, legal form standardization (ISO
20275), and role classification.
The local legal form (Cégforma) is extracted deterministically from the TRE and then enriched with AI for ISO 20275 code and standardized company type.
Legal representatives are extracted from Section 13 (§13) of the Trade Register Extract: “A vezető tisztségviselő(k), a képviseletre jogosult(ak) adatai” (Data of executive officers and those authorized to represent).
Individual representatives: Name, birth date, residence address (Hungarian or foreign), role, start date, tax ID
Company representatives: Legal name, registered address, role, start date
Representation mode: Extracted as “együttes” (joint) or “önálló” (individual)
Legal representatives are only available for active companies. Closed
companies use a minimal TRE format that may not include Section 13. This is a
register limitation.
The following roles are deterministically mapped from Hungarian registry terminology — no AI is involved for these known roles:
Hungarian Role
English Translation
Standardized Role
ügyvezető
Managing Director
Managing Director
cégvezető
Company Manager
Managing Director
vezető tisztségviselő
Executive Officer
Managing Director
ügyvezető (vezető tisztségviselő)
Managing Director (Executive Officer)
Managing Director
vezérigazgató
CEO
CEO
igazgatósági tag
Board Member
Board Member
elnök
Chairman
Chairman
képviseletre jogosult tag
Member with Representation Rights
Other
Roles matching the table above are mapped deterministically (AI is
skipped). For roles not matching any known pattern, AI provides the English
translation and standardized classification. The original Hungarian role name
is always preserved as localName.
Hungary uses the TEÁOR (Tevékenységek Egységes Ágazati Osztályozási Rendszere) classification, which is the Hungarian adaptation of NACE Rev. 2. Activity codes are extracted from the TRE and mapped across three levels:
TEÁOR codes are extracted directly from the TRE and are never AI-inferred.
The main activity is identified when the TRE marks it as “Főtevékenység” (main
activity).
Business premises (§6) and branch offices (§7) from TRE
Legend: ✅ Available | ❌ Not Available | ⚠️ Conditionally Available | Async = Asynchronous processing
Sole Entrepreneurs (EV) are not registered in the Cégjegyzék (Company
Register) managed by OCCSZ. They are registered in a separate system (Egyéni
Vállalkozói Nyilvántartás) that is not covered by this integration. Egyéni
Cég (EC) — Individual Company — is a different entity type that is
registered in the Cégjegyzék and falls under the Commercial Companies column.
Document prices are fetched in real-time from the OCCSZ API and displayed
in the price field before purchase. Pricing is based on document size
(kbyte-based billing) and may vary between companies.
TRE price is variable. Unlike flat-rate documents, the Trade Register
Extract price depends on the document size. Larger companies with more
historical data will have higher prices. The price is always shown before
purchase.
Company profile pricing for Hungary is variable — the total cost depends on the underlying Trade Register Extract price (which varies by company) plus a fixed profile fee.
Budget control. Because pricing is variable, the profileMaxBudget
parameter lets you set a maximum spend (in credit cents). If the estimated
cost exceeds your budget, the request fails with a budget_exceeded error
that includes the quoted price, so you can retry with a higher budget.
For a given entity, data follows a deterministic retrieval path:
Company Profile Resolution Flow:
Search — Query the OCCSZ XML API by registration number to verify the company exists and obtain the tax number
Available Documents — Fetch the TRE price quote from OCCSZ
Budget check — If profileMaxBudget is set, compare the estimated cost (TRE with markup + fixed profile fee) against the budget. Reject with budget_exceeded if over budget
Download TRE — Purchase and download the Trade Register Extract PDF from OCCSZ
PDF → Markdown — Convert the PDF to structured markdown using PyMuPDF4LLM
Deterministic parsing — Parse the markdown to extract all structured data (name, address, legal form, capital, representatives, shareholders, activities, establishments, status)
AI enrichment (parallel) — Enrich in parallel: address parsing, legal form standardization (ISO 20275), currency parsing, role classification for legal representatives, establishment address parsing
VIES VAT validation — A qualified VIES check validates the VAT number (HU + first 8 digits of tax number) and obtains a consultation number
Document Resolution Flow:The Trade Register Extract PDF is purchased directly from the OCCSZ API. The price is fetched first, then the document is downloaded and stored. The same PDF used for company profile extraction is delivered as the document.
Note: Closed companies have active: false and may have empty legal representatives. The closureDate is extracted from the TRE. Companies that were in liquidation (“végelszámolás alatt”) before deletion retain that marker in their legal name.
Available Documents
Documents are returned when "dataPoints": ["availableDocuments"] is requested.
API Category
Document
Notes
tradeRegisterExtract
Cégkivonat (Trade Register Extract) PDF
Dynamic pricing from OCCSZ. Available for all companies (active and closed)
Both Cégjegyzékszám and Adószám may contain hyphens (e.g., 01-09-430929, 28383040-2-42). These are stripped internally before processing.
Leading zeros in registration numbers
Registration numbers always have 10 digits. Numbers with leading zeros (e.g., 0109430929 — Budapest court code 01) are preserved correctly.
Liquidation detected from company name
Companies in liquidation are identified by the presence of “végelszámolás alatt” or “v.a.” in the legal name, not from a dedicated status field.
TRE price is variable
Trade Register Extract pricing is kbyte-based. Larger companies (more history, more sections) cost more. Prices are always shown before purchase.
Company profile = TRE
The company profile is entirely derived from the Trade Register Extract PDF. Requesting a company profile will purchase and parse the TRE.
No shareholders for all company types
Part II (ownership section) is not present in all TRE documents. Availability depends on the company type and registry content.
No legal reps for closed companies
Section 13 (legal representatives) may be absent in TRE documents for closed companies.
Closure reason defaults to OTHER
When a company is closed (“Törölve”) and the deletion method doesn’t match a known pattern, the closure reason defaults to OTHER. Known patterns: átalakulás (Merger), felszámolás (Bankruptcy), kényszertörlés (Administrative Dissolution), végelszámolás (Voluntary Dissolution).
Tax number truncated to 13 digits
The tax number (Adószám) is truncated to 13 characters in the identifiers to standardize the format.
TEÁOR = NACE
Hungarian TEÁOR codes map 1:1 to NACE Rev. 2 codes (same numeric values, different formatting).
HUF to EUR conversion
Document prices are quoted in HUF by the OCCSZ API and converted to EUR for the API response.