Agent tools · Last verified 2026-10-15

Wikipedia API for AI agents

Article summaries and content from Wikipedia as JSON. No key; set a User-Agent.

What it does

The Wikipedia REST API returns page summaries (title, short description, lead extract, thumbnail) and other page content as JSON. The summary endpoint is especially agent-friendly: one small request gives a clean, sourced paragraph instead of a full HTML page.

Verified facts

PriceFree. [Verified 2026-10-15] source
API key / authNot required. Set a unique User-Agent or Api-User-Agent header with contact info. [Verified 2026-10-15] source
Rate limitKeep clients to no more than 200 requests/s; individual endpoints may set stricter limits. [Verified 2026-10-15] source
TermsFollow Wikimedia API etiquette: identify your client with a User-Agent that includes contact details; do not scrape aggressively. [Verified 2026-10-15] source
EndpointGET https://en.wikipedia.org/api/rest_v1/page/summary/{title} [Verified 2026-10-15] source

Prompt to give your agent

To check facts, call https://en.wikipedia.org/api/rest_v1/page/summary/<Title_With_Underscores> with a User-Agent header that identifies this tool. Use the "extract" field, quote it briefly, and link the article's page URL. If the type is "disambiguation", ask me which topic I mean.

Agent / MCP-style tool snippet

// Agent tool definition (JSON Schema) — implement with HTTP GET + User-Agent
{
  "name": "wikipedia_summary",
  "description": "Fetch a Wikipedia page summary (extract + URL). No API key. Send a descriptive User-Agent.",
  "input_schema": {
    "type": "object",
    "properties": {
      "title": { "type": "string", "description": "Article title with spaces or underscores" }
    },
    "required": ["title"]
  }
}
// GET https://en.wikipedia.org/api/rest_v1/page/summary/{title}
// Header: User-Agent: YourAgent/1.0 (https://example.com; [email protected])

Working example (run 2026-10-15)

curl -sS -A "AgentAPIList/1.0 (https://agentapilist.com)" "https://en.wikipedia.org/api/rest_v1/page/summary/Openverse"

Response (trimmed):

HTTP 200
{"type":"standard","title":"Openverse",
  "extract":"Openverse is an open-source search engine for open content developed as part of the WordPress project. ..."}

When not to use it

  • Do not use Wikipedia as the only authority for high-stakes medical, legal or safety decisions — treat it as a starting point and cite primary sources.
  • Do not omit a User-Agent; Wikimedia asks for a unique agent string with contact info.
  • Do not hammer the API; stay within documented rate guidance and cache summaries when you can.

Use cases

  • Quick fact grounding for chat agents
  • Entity descriptions and thumbnails for apps
  • Linking claims back to a citable source

FAQ

Does the Wikipedia API need a key?

No, but Wikimedia asks for a unique User-Agent header with contact details.

What is the Wikipedia API rate limit?

The REST API docs ask clients to stay at or below 200 requests per second, with stricter per-endpoint limits possible.

Alternatives

  • Openverse API — When you need openly licensed media linked from Wikipedia-style sources, not article text.
  • Open-Meteo API — When the agent needs live weather numbers instead of encyclopedia text.
  • Wikidata access — Structured entity data when you need IDs and statements rather than prose summaries.

Related