Wikipedia Article Extractor

Turn a list of titles — or a single search query — into clean, structured Wikipedia articles. For every article you get a flat JSON row with the summary, short description, canonical URL, thumbnail, language, page type and (when available) geo-coordinates, plus the full article text on demand.

Built on Wikipedia's free public REST API and Action API — no API key, no login and no proxy required.

Why use it

• Ready-to-use data. Each article is flattened into predictable fields, so it drops straight into a spreadsheet, database or downstream model.
• Titles or search — or both. Feed exact titles, a search query, or a mix; results are merged and de-duplicated.
• Any language edition. Point it at en, de, fr, es, or any other Wikipedia edition.
• Optional full text. Flip one switch to also pull the complete plain-text body (truncated to ~5000 characters).
• Robust by design. Redirects are followed, missing pages are handled gracefully, and one bad title never fails the whole run — it is recorded with an error field instead.

Pricing

This actor runs on a pay-per-event model: you are charged once per successfully extracted article (article-scraped). Rows that fail (for example, a title that does not exist) are written to the dataset but are never charged.

Input

Field	Type	Default	Description
titles	array	["Web scraping", "Instagram"]	Exact Wikipedia article titles to extract. Redirects are followed automatically.
searchQuery	string	—	Optional. Search Wikipedia and extract the matching articles.
language	string	en	Wikipedia language edition code.
fullText	boolean	false	Also extract the full plain-text body (adds one request per article).
maxResults	integer	20	Maximum number of articles to output (titles + search results combined).

Provide at least one of titles or searchQuery.

Example input

{
  "titles": ["Web scraping", "Instagram"],
  "language": "en",
  "fullText": false,
  "maxResults": 20
}

Or search instead:

{
  "searchQuery": "web scraping",
  "maxResults": 5
}

Output

Each dataset record looks like this:

{
  "title": "Web scraping",
  "description": "Data scraping used for extracting data from websites",
  "extract": "Web scraping, web harvesting, or web data extraction is data scraping used for extracting data from websites…",
  "url": "https://en.wikipedia.org/wiki/Web_scraping",
  "thumbnail": null,
  "lang": "en",
  "type": "standard",
  "coordinates": null
}

Field	Description
title	The resolved article title.
description	Short one-line description.
extract	Plain-text summary of the article.
url	Canonical desktop URL of the article.
thumbnail	Thumbnail image URL, or null.
lang	Language edition code.
type	Page type (standard, disambiguation, etc.).
coordinates	{ lat, lon } for geo-located articles, or null.
fullText	Full plain-text body (only when fullText is enabled).

If an article can't be extracted (e.g. the page doesn't exist), the row is { "title": "<title>", "error": "<message>" } instead.

Data source

This actor uses Wikipedia's official public APIs:

• Summary (REST v1): https://<lang>.wikipedia.org/api/rest_v1/page/summary/<title>
• Full text & search (Action API): https://<lang>.wikipedia.org/w/api.php

The APIs are free and require no authentication. Per Wikipedia's API etiquette, every request sends a descriptive User-Agent header.

License

MIT © Assia Fadli