Wikipedia & Wikidata Company Data — LEI, ISIN, Ticker

Turn a company name — or an entire Wikipedia category — into a structured record built around the identifiers that actually let you join it to other datasets: Wikidata QID, LEI, ISIN, stock ticker and official website, plus industry, country, headquarters, founding date, employees, parent company and founders. Optional infobox financials and monthly pageviews.

• 🔗 Built for matching, not just reading — LEI 68%, ISIN 73%, ticker 92%, official website 96% on a 598-company NYSE sweep (measured)
• 📚 Category mode = real volume — point it at "Companies listed on the New York Stock Exchange" and get 598 companies in 68 seconds, not the handful you could type by hand
• 🏷️ Readable, not raw — referenced Wikidata entities are resolved to labels ("United States", not "Q30")
• 🆓 No key, no proxy — official Wikipedia, Wikidata and Wikimedia endpoints, batched and polite

Quick start (API)

Four companies with identifiers, infobox financials and a year of pageviews:

curl -X POST "https://api.apify.com/v2/acts/foxlabs~wikipedia-company-scraper/run-sync-get-dataset-items?token=YOUR_APIFY_TOKEN" \
  -H "Content-Type: application/json" \
  -d '{ "mode": "lookup", "companies": ["Apple Inc.", "Shopify", "Siemens"], "includeInfobox": true, "pageviewMonths": 12 }'

Prefer no code? Open the Input tab, keep the prefilled companies, and press Start.

What you get

Field	Type	Description
companyName / wikipediaUrl	string	Article title and URL
summary	string	Lead paragraph of the article, plain text
wikidataQid / wikidataUrl	string	The Wikidata entity — your join key to the wider open-data graph
officialWebsite	string	Company website as recorded on Wikidata
lei	string	Legal Entity Identifier — the global regulatory ID
isin	string	Security identifier for the listed entity
tickerSymbol / stockExchange	string	Ticker and where it trades
industry / country / headquarters	string	Resolved to readable labels
inception	string	Founding date — 2006, 1866-03 or 1976-04-01 depending on how precisely Wikidata records it
employees	number	Employee count where Wikidata has one
parentOrganization / subsidiaries / founders	string / array	Corporate relations, as labels
legalForm	string	e.g. public company, GmbH
pageId / language / license / fetchedAt	string	Provenance, including the CC BY-SA attribution
infobox	object	Optional: revenue, operating & net income, total assets, key people, products from the article infobox
pageviewsMonthly / pageviewsTotal / pageviewsLatest	array / number	Optional: monthly Wikipedia views — a public-attention signal

Sample output

A real record (2026-07-31, trimmed):

{
  "companyName": "Apple Inc.",
  "wikipediaUrl": "https://en.wikipedia.org/wiki/Apple_Inc.",
  "wikidataQid": "Q312",
  "officialWebsite": "https://www.apple.com/",
  "lei": "HWUPKR0MPOU8FGXBT394",
  "isin": "US0378331005",
  "tickerSymbol": "AAPL",
  "industry": "software industry",
  "country": "United States",
  "inception": "1976-04-01",
  "employees": 115000,
  "founders": ["Steve Wozniak", "Ronald Wayne", "Steve Jobs"],
  "infobox": { "revenue": "US$416 billion (2025)" },
  "pageviewsLatest": 171714,
  "license": "Text from Wikipedia/Wikidata, CC BY-SA 4.0"
}

Download as JSON, CSV, Excel or HTML, or read it from the API.

Modes & options

• Look up companies — paste names as they appear on Wikipedia (or article URLs). Resolved in batches, so long lists stay quick.
• Browse a category — give one or more Wikipedia categories and the Actor walks them with continuation until the category is exhausted or your maxResults is hit. categoryDepth optionally follows subcategories.
• Wikidata identifiers — on by default; this is what makes the output joinable.
• Infobox financials — optional, one extra fetch per company; adds revenue and income figures Wikidata usually lacks.
• Pageview history — optional, 1-60 months of monthly views per company.

Example inputs (copy & paste)

// 1) A short list, fully enriched
{ "mode": "lookup", "companies": ["Nestlé", "Siemens", "Toyota"], "includeInfobox": true }

// 2) Every NYSE-listed company Wikipedia knows
{ "mode": "category", "categories": ["Companies listed on the New York Stock Exchange"], "maxResults": 0 }

// 3) US software companies
{ "mode": "category", "categories": ["Software companies of the United States"], "maxResults": 1000 }

// 4) Attention trend for a brand set
{ "mode": "lookup", "companies": ["Shopify", "Etsy", "BigCommerce"], "pageviewMonths": 24 }

// 5) Deep sweep including subcategories
{ "mode": "category", "categories": ["Banks of Germany"], "categoryDepth": 1, "maxResults": 0 }

Use cases

• Entity resolution & data enrichment. You have company names; you need something to match on. LEI, ISIN, ticker and the official domain are exactly the keys that link a name to registry, market and ownership datasets.
• Building a company reference list. Category mode gives you a clean, deduplicated universe (an index, a sector, a country) with identifiers attached.
• Market and competitor research. Industry, country, headquarters, founding date, employees, parent and subsidiaries in one flat row.
• Brand attention tracking. Monthly pageviews are a free, long-running proxy for public interest — schedule it and trend the series.
• Knowledge-graph seeding. Every row carries its Wikidata QID, so you can expand into the wider graph whenever you need more.

Performance & cost

Official endpoints only, no proxy, no key, everything batched: articles 20 at a time (the API's own extract limit), Wikidata entities 50 at a time. Measured on the platform (2026-08-01): a 598-company category sweep took 68 seconds; a 4-company lookup with infobox and 12 months of pageviews took 6 seconds. Each company appears once — when a category lists the same article under two titles, the rows are de-duplicated on the resolved page and the run reports how many it removed. Optional layers cost one extra request per company, so leave them off for big sweeps unless you need them.

Integrations

JavaScript (apify-client):

import { ApifyClient } from 'apify-client';
const client = new ApifyClient({ token: 'YOUR_APIFY_TOKEN' });
const run = await client.actor('foxlabs/wikipedia-company-scraper').call({
  mode: 'category', categories: ['Companies listed on the New York Stock Exchange'], maxResults: 0,
});
const { items } = await client.dataset(run.defaultDatasetId).listItems();

Python (apify-client):

from apify_client import ApifyClient
client = ApifyClient("YOUR_APIFY_TOKEN")
run = client.actor("foxlabs/wikipedia-company-scraper").call(run_input={
    "mode": "lookup", "companies": ["Apple Inc.", "Siemens"], "includeInfobox": True,
})
for item in client.dataset(run["defaultDatasetId"]).iterate_items():
    print(item["companyName"], item.get("lei"), item.get("isin"))

Also works with Make / n8n / Zapier, scheduled runs, webhooks and the Apify MCP server.

Data quality (measured 2026-07-31)

Category sweep, "Companies listed on the New York Stock Exchange", n = 598 (every row a distinct company — verified 0 duplicate pages):

Field	Fill
Company name, summary	100%
Wikidata QID	99.8%
Official website	96%
Headquarters	94%
Ticker symbol	92%
Founding date	91%
Country	89%
Industry	76%
ISIN	73%
LEI	68%
Employees	21%

Read those honestly: identifiers exist where Wikidata's editors recorded them. Big listed companies are well covered; employee counts are the weakest field at 21% because few company entities carry one. Nothing is inferred or guessed — an empty field means the source is empty. Every row carries its license string, because Wikipedia and Wikidata text is CC BY-SA 4.0 and attribution travels with the data.

Joining this data to the rest of your stack

The identifier fields exist so the output slots straight into other datasets — that's the whole point:

You have	Join on	Where it goes next
lei	Legal Entity Identifier	LEI Company Ownership for the registered legal entity, parent chain and registration status
isin / tickerSymbol	Security identifiers	Any market-data source; SEC EDGAR Financials for US filers
officialWebsite → domain	The most reliable cross-source key	Company Data Enrichment and Owler Intelligence — match on domain rather than name, which is where most enrichment goes wrong
wikidataQid	Wikidata entity	The wider open-data graph — every other identifier Wikidata holds for that company
companyName + country	Registry lookup	The foXLabs registry suite (UK, France, India, Estonia, …)

A practical recipe: run category mode over your sector, keep the rows that carry an lei, and use those as the spine of your company list — LEI is a regulator-issued identifier, so it survives renames and acquisitions in a way company names do not.

Pricing

Pay per result — billed per company record, with an Apify free tier to evaluate. The underlying endpoints are free, so there are no proxy or third-party costs.

FAQ

Where does the data come from? The official Wikipedia action API, Wikidata and the Wikimedia pageviews API. No key, no login.

Why Wikidata rather than just the infobox? Infoboxes are prose in a table; Wikidata is structured and carries the identifiers (LEI, ISIN, ticker, QID) that let you join records across datasets. Infobox parsing is still available for revenue and income figures Wikidata rarely stores.

Can I get every company in an industry or country? That's what category mode is for. Wikipedia's category system is the index — one category returned 598 companies in the measured run, and continuation keeps going until the category is exhausted.

Why is employees often empty? Because most Wikidata company entities don't record it (21% in the measured sweep). The infobox layer sometimes fills the gap.

What does inception look like? Whatever precision Wikidata holds: 1866, 1866-03 or 1976-04-01. Year-only dates are returned as a year rather than a fake full date.

Is this legal / can I reuse it? Wikipedia and Wikidata content is published under CC BY-SA 4.0 (Wikidata's structured data is CC0). Attribution is required for Wikipedia text — each row carries a license field to make that easy.

Does it work for non-English companies? Yes, if the company has an English Wikipedia article; the Wikidata layer is language-independent. Use the English title (e.g. "Nestlé", "Volkswagen Group").

Troubleshooting

• A company came back empty / was skipped → the title didn't resolve to an English Wikipedia article. Check the exact spelling on Wikipedia (the run log names every skipped title), or pass the full article URL.
• Fewer rows than the category has → raise maxResults (0 = unlimited).
• No infobox object → that article has no infobox, or its rows use labels this Actor doesn't map; infoboxStatus says which.
• No pageviews → very new or rarely visited articles can have no data for the window.

Notes, limits & legal

• Content belongs to Wikipedia/Wikidata contributors and is CC BY-SA 4.0 (structured Wikidata claims: CC0). Attribution is included in every row; keep it when you redistribute.
• Coverage mirrors Wikipedia's own coverage: large, listed and notable companies are well described; small private firms are usually absent entirely.
• The Actor sends a descriptive User-Agent with contact details, as Wikimedia's API etiquette asks.

Support

Questions, a field you'd like added, or a custom build? Open the Issues tab, or e-mail info@foxlabs.com.tr. We reply fast.

If this Actor saves you time, a ⭐ review really helps.

Changelog

0.2 — 2026-07-31

• Wikidata identifiers (QID, LEI, ISIN, ticker, exchange, website) plus industry, country, HQ, inception, employees, parent, subsidiaries and founders, with referenced entities resolved to readable labels.
• Category mode — walk an entire Wikipedia category with continuation (measured: 598 companies in 68 seconds).
• Optional layers — infobox financials and monthly pageviews.
• Three data bugs caught by pre-release testing: ticker symbol lives as a qualifier on the stock-exchange statement, not as its own claim (fill 0% → 92%); year-only dates were emitted as invalid 1866-00-00 and now respect Wikidata's precision flag; and the article API caps extracts at 20 pages per request regardless of batch size, which had left summary ~40% filled — batching at 20 took it to 100%.

0.1 — 2026-05-20

• Initial release: Wikipedia infobox HTML parsing for a supplied list of articles.

---

Part of the foXLabs data platform — official public-data company, ownership, registry & market intelligence scrapers. Browse the full suite at data.foxlabs.com.tr.