EU Legal Entity Extractor - VAT, Registry & Imprint API

What does EU Legal Entity Extractor do?

EU Legal Entity Extractor collects company names and legal identifiers published on EU/UK websites, with source-linked contact and imprint evidence. Try one public imprint page in the Actor, then export the result or connect it through Apify’s API and integrations. Optional platform scheduling can repeat a check, but the Actor does not create a monitor or compare runs for you.

Why use EU Legal Entity Extractor?

Use it when a vendor-onboarding or CRM-review queue needs website-published identity evidence before a person checks an official register. Country and confidence are inferred signals, so they should not determine a KYC/KYB decision by themselves.

How to use EU Legal Entity Extractor

1. Supply one public legal-page URL in the Input tab.
2. Keep evidence enabled and budget for the $0.002 attempt plus possible $0.008 found-entity event.
3. Review the returned identifiers against their sources before expanding the input.

Input

This public-site example shows the input format, not a promised result:

{
  "startUrls": [{ "url": "https://www.bosch.de/impressum/" }],
  "maxPagesPerSite": 4,
  "concurrency": 1,
  "requestTimeoutSecs": 10,
  "includeEvidence": true,
  "validateVat": false
}

Use startUrls or bare domains, with at most 100 combined entries including duplicates. The Input tab describes the remaining options and their limits.

Output

This shortened illustration uses a placeholder company, not a live extraction or customer result:

{
  "domain": "company.example",
  "status": "PARTIAL",
  "companyName": "Example GmbH",
  "vatIds": [],
  "registryIds": [],
  "legalPageUrls": ["https://company.example/impressum"],
  "vatValidations": []
}

You can download the dataset in various formats such as JSON, HTML, CSV, or Excel. Retain retrievedAt and source evidence with the exported values so a reviewer can revisit the publication context.

Data table

Field	Meaning
companyName, country	Extracted name and inferred jurisdiction, not registry-confirmed identity.
vatIds, registryIds	Normalized VAT strings and matched register identifiers.
vatValidations	Optional VIES results, including unavailable or unsupported states.
evidence, sourceUrls	Snippets and checked source URLs for manual review.
status, errors	Detection status and source or budget problems.

Pricing / Cost estimation

How much does it cost to check a website? The current fees are $0.002 per attempted website (website-processed) plus $0.008 for a qualifying FOUND entity (legal-entity-found), totalling $0.01 when both events apply. The found-entity event requires FOUND status with at least one VAT or registry identifier; PARTIAL, NO_LEGAL_ENTITY and ordinary FETCH_FAILED rows incur no found-entity event, although the attempt fee can apply.

The run reserves both fees before starting a website, even when the eventual result needs only the attempt fee. Check the Pricing tab for current platform terms and plan limits, and inspect RUN_SUMMARY before repeating an uncertain run. No additional private VAT-service key is required.

For stopped runs, check RUN_ABORT_STATE; an incomplete checkpoint or unknown final counts take precedence over any stale RUN_SUMMARY. Confirmed-so-far counts are lower bounds, not final totals.

Tips and advanced options

Start at the known imprint or legal page to make the limited check more relevant. Each site has a 20-second deadline, a 1 MiB downloaded-response budget and a 4 MiB decoded-text budget; failed requests, redirects and optional VIES requests consume its attempt allowance. maxPagesPerSite sets that allowance, not a guarantee that every candidate page will be read.

validateVat defaults to true; set it false to skip VIES. When enabled, up to five detected IDs may be checked within the remaining site budget. A service outage or incomplete check is not an invalid-VAT verdict, and non-supported identifiers cannot be validated through VIES.

FAQ, disclaimers and support

The Actor reads static public pages without a browser or OCR, so script-rendered or image-only details may be missed. It does not bypass sign-in, configure proxy rotation or verify corporate ownership, sanctions or legal compliance. A missing identifier means it was not detected in the checked pages, rather than proving that none exists.

Only inspect websites you are permitted to check and verify important identifiers independently. Use the Issues tab with a sanitized example for support or to discuss a custom scope.