Wikipedia / MediaWiki

aid: wikipedia
name: Wikipedia / MediaWiki
description: >-
  Wikipedia is the free, multilingual online encyclopedia operated by the non-profit
  Wikimedia Foundation. The platform exposes its content and structured data through
  several public APIs: the original MediaWiki Action API (action=query|edit|parse|upload),
  the MediaWiki Core REST API (page CRUD, search, transforms), the Wikimedia REST API
  (page summaries, mobile HTML, math, citation, reading lists, recommendations), the
  Wikidata Query Service SPARQL endpoint (structured knowledge graph queries), and the
  Wikimedia Enterprise APIs (commercial Snapshot, On-demand, Realtime). All free APIs are
  governed by the Wikimedia API usage guidelines: contactable User-Agent, serial requests,
  maxlag for bots.
url: https://api.wikimedia.org/wiki/Main_Page
image: https://upload.wikimedia.org/wikipedia/commons/thumb/8/80/Wikipedia-logo-v2.svg/120px-Wikipedia-logo-v2.svg.png
specificationVersion: '0.20'
created: '2026-05-28'
modified: '2026-05-29'
x-source: public-apis/public-apis
x-type: opensource
x-category: Open Data
x-tier: 1
x-tier-reason: Most prominent open-knowledge API in the world; non-profit Wikimedia Foundation
tags:
  - Open Data
  - Public APIs
  - Open Knowledge
  - Encyclopedia
  - Knowledge Graph
  - Open Source
  - Non-Profit
apis:
  - name: MediaWiki Action API
    description: >-
      The original programmatic interface to MediaWiki, exposed at /w/api.php on every
      MediaWiki installation. Operations are dispatched via the action= query parameter
      (query, parse, edit, upload, login, patrol, ...). JSON is the recommended format.
      Primary write surface for Wikimedia projects.
    humanURL: https://www.mediawiki.org/wiki/API:Main_page
    baseURL: https://en.wikipedia.org/w/api.php
    tags:
      - Action API
      - Read
      - Write
      - Edit
      - Search
    properties:
      - type: Documentation
        url: https://www.mediawiki.org/wiki/API:Main_page
      - type: Sandbox
        url: https://en.wikipedia.org/wiki/Special:ApiSandbox
      - type: OpenAPI
        url: openapi/wikipedia-mediawiki-action-api-openapi.yaml
      - type: Tutorials
        url: https://www.mediawiki.org/wiki/API:Tutorial
      - type: Etiquette
        url: https://www.mediawiki.org/wiki/API:Etiquette

  - name: MediaWiki Core REST API
    description: >-
      Modern REST surface on every MediaWiki installation under /w/rest.php/v1/. Provides
      page CRUD, full-text search, file metadata, history, individual revision retrieval,
      revision comparison, and wikitext <-> HTML transforms. Mirrored at
      api.wikimedia.org/core/v1/wikipedia/{lang}/.
    humanURL: https://www.mediawiki.org/wiki/API:REST_API
    baseURL: https://en.wikipedia.org/w/rest.php/v1
    tags:
      - REST API
      - Core
      - Pages
      - Search
      - Transforms
    properties:
      - type: Documentation
        url: https://www.mediawiki.org/wiki/API:REST_API/Reference
      - type: OpenAPI
        url: openapi/wikipedia-mediawiki-core-rest-openapi.yaml
      - type: APIPortal
        url: https://api.wikimedia.org/wiki/Core_REST_API

  - name: Wikimedia REST API v1
    description: >-
      Caching-optimized REST API for high-volume, anonymous read traffic. Provides page
      summaries, mobile-tailored HTML, media lists, citation extraction, math rendering,
      content transforms, page recommendations, and reading lists. Limit clients to no more
      than 200 requests/s and provide a contactable User-Agent.
    humanURL: https://en.wikipedia.org/api/rest_v1/
    baseURL: https://en.wikipedia.org/api/rest_v1
    tags:
      - REST API
      - Cached
      - Summaries
      - Mobile
      - Citation
      - Math
    properties:
      - type: Documentation
        url: https://en.wikipedia.org/api/rest_v1/
      - type: OpenAPI
        url: openapi/wikipedia-rest-v1-openapi.yaml
      - type: Specification
        url: https://en.wikipedia.org/api/rest_v1/?spec

  - name: Wikidata Query Service (SPARQL)
    description: >-
      SPARQL 1.1 endpoint for Wikidata - the structured knowledge graph behind Wikipedia
      infoboxes, Commons captions, and 100M+ items. Single endpoint /sparql accepts SELECT,
      ASK, CONSTRUCT, DESCRIBE queries. 60s query timeout, 5 concurrent queries per IP. Has
      a 'main' / 'scholarly' graph split available at separate hostnames for heavy workloads.
    humanURL: https://query.wikidata.org
    baseURL: https://query.wikidata.org
    tags:
      - SPARQL
      - Knowledge Graph
      - Linked Data
      - Wikidata
    properties:
      - type: Documentation
        url: https://www.wikidata.org/wiki/Wikidata:SPARQL_query_service
      - type: OpenAPI
        url: openapi/wikipedia-wikidata-sparql-openapi.yaml
      - type: QueryHelp
        url: https://www.wikidata.org/wiki/Wikidata:SPARQL_query_service/Wikidata_Query_Help
      - type: Editor
        url: https://query.wikidata.org

  - name: Wikimedia Enterprise API
    description: >-
      Commercial, SLA-backed Wikimedia data feed run by Wikimedia LLC (the Foundation's
      subsidiary). Provides Snapshot (bulk twice-monthly or daily), On-demand (per-article
      structured contents), and Realtime (streaming updates) APIs. Free developer tier
      includes 5,000 on-demand requests/month and 15 snapshot/month. Paid plans add
      unlimited usage and realtime streams.
    humanURL: https://enterprise.wikimedia.com/
    baseURL: https://api.enterprise.wikimedia.com
    tags:
      - Enterprise
      - Commercial
      - Snapshot
      - On-Demand
      - Realtime
      - Structured Contents
    properties:
      - type: Documentation
        url: https://enterprise.wikimedia.com/docs/
      - type: OpenAPI
        url: openapi/wikipedia-wikimedia-enterprise-openapi.yaml
      - type: DataDictionary
        url: https://enterprise.wikimedia.com/docs/data-dictionary/
      - type: Pricing
        url: https://enterprise.wikimedia.com/pricing/
      - type: SDK
        title: Python SDK
        url: https://github.com/wikimedia-enterprise/wme-sdk-python
      - type: SDK
        title: Go SDK
        url: https://github.com/wikimedia-enterprise/wme-sdk-go
      - type: Postman
        url: https://github.com/wikimedia-enterprise/postman

common:
  - type: Website
    url: https://www.wikipedia.org
  - type: APIPortal
    url: https://api.wikimedia.org/wiki/Main_Page
  - type: APICatalog
    url: https://api.wikimedia.org/wiki/API_catalog
  - type: Documentation
    url: https://www.mediawiki.org/wiki/API:Main_page
  - type: Foundation
    url: https://wikimediafoundation.org/
  - type: Governance
    url: https://meta.wikimedia.org/wiki/Wikimedia_Foundation
  - type: License
    url: https://creativecommons.org/licenses/by-sa/4.0/
    title: CC BY-SA 4.0 (article content)
  - type: License
    url: https://creativecommons.org/publicdomain/zero/1.0/
    title: CC0 1.0 (Wikidata)
  - type: Policy
    url: https://foundation.wikimedia.org/wiki/Policy:Wikimedia_Foundation_API_Usage_Guidelines
    title: API Usage Guidelines
  - type: PublicAPIsListing
    url: https://github.com/public-apis/public-apis
  - type: GitHubOrganization
    url: https://github.com/wikimedia
  - type: GitHubOrganization
    url: https://github.com/wikimedia-enterprise
    title: Wikimedia Enterprise GitHub Org
  - type: GitHubOrganization
    url: https://github.com/wmde
    title: Wikimedia Deutschland (Wikidata) GitHub Org
  - type: SourceCode
    url: https://gerrit.wikimedia.org/
    title: Wikimedia Gerrit (canonical source)
  - type: Status
    url: https://www.wikimediastatus.net/
  - type: BulkDownload
    url: https://dumps.wikimedia.org/
    title: Wikimedia Database Dumps
  - type: Stream
    url: https://stream.wikimedia.org/v2/stream
    title: EventStreams (SSE)
  - type: Tools
    title: MCP Server (Wikipedia, Python, Rudra-ravi)
    url: https://github.com/Rudra-ravi/wikipedia-mcp
  - type: Tools
    title: MCP Server (Wikipedia, TypeScript, timjuenemann)
    url: https://github.com/timjuenemann/wikipedia-mcp
  - type: Tools
    title: MCP Server (Wikipedia, .NET, ajayindfw)
    url: https://github.com/ajayindfw/WikipediaMcpServer
  - type: Tools
    title: MCP Server (Wikipedia, Vercel HTTP, Ravishka17)
    url: https://github.com/Ravishka17/Wikipedia-MCP
  - type: Tools
    title: MCP Server (Wikipedia caching+batch, 1999AZZAR)
    url: https://github.com/1999AZZAR/wikipedia-mcp-server
  - type: Tools
    title: MCP Server (MediaWiki multi-wiki, shiquda)
    url: https://github.com/shiquda/mediawiki-mcp-server
  - type: Tools
    title: MCP Server (MediaWiki search+edit, olgasafonova)
    url: https://github.com/olgasafonova/mediawiki-mcp-server
  - type: Tools
    title: MCP Server (MediaWiki authoring, mbruton)
    url: https://github.com/mbruton/mediawiki-mcp
  - type: Tools
    title: MCP Server (Wikidata, Wikimedia Deutschland)
    url: https://github.com/wmde/WikidataMCP
  - type: Tools
    title: MCP Server (Wikidata SPARQL, cyanheads)
    url: https://github.com/cyanheads/wikidata-mcp-server
  - type: Plans
    url: plans/wikipedia-plans-pricing.yml
  - type: RateLimits
    url: rate-limits/wikipedia-rate-limits.yml
  - type: FinOps
    url: finops/wikipedia-finops.yml
  - type: NaftikoCapability
    url: capabilities/wikipedia-knowledge-research.yaml
  - type: NaftikoCapability
    url: capabilities/wikipedia-content-authoring.yaml
  - type: NaftikoCapability
    url: capabilities/wikipedia-knowledge-graph.yaml
  - type: NaftikoCapability
    url: capabilities/wikipedia-enterprise-snapshot.yaml
  - type: Rules
    url: rules/wikipedia-spectral-rules.yml
  - type: Vocabulary
    url: vocabulary/wikipedia-vocabulary.yaml
  - type: JsonLD
    url: json-ld/wikipedia-context.jsonld

features:
  - name: 300+ Languages
    description: Wikipedia exists in over 300 languages; every language edition exposes the same API surfaces under {lang}.wikipedia.org.
  - name: Page Summaries
    description: Short, plain-text extracts plus thumbnail and Wikidata QID via the /page/summary/{title} endpoint - widely used by mobile clients, voice assistants, and search engines.
  - name: Wikitext To HTML Transforms
    description: Convert wikitext to rendered HTML (and back) on-demand via /transform/wikitext/to/html/{title} - the same engine that powers VisualEditor.
  - name: SPARQL Over A Public Knowledge Graph
    description: Query 100M+ Wikidata items with full SPARQL 1.1 including federated SERVICE clauses, label resolution, and geo queries.
  - name: Real-Time Edit Stream
    description: Server-Sent Events stream at stream.wikimedia.org/v2/stream emits every edit, log, and revision-create across all projects.
  - name: Bulk Database Dumps
    description: Full periodic XML and SQL dumps of every Wikimedia project at dumps.wikimedia.org - the canonical bulk-research format.
  - name: Cached Anonymous Reads At 200 RPS
    description: The REST API v1 is heavily Varnish-cached, allowing anonymous reads up to ~200 requests/s per client with a contactable User-Agent.
  - name: OAuth 2.0 Write Access
    description: Modern write access (page edits, file uploads) is authenticated via OAuth 2.0 bearer tokens issued at meta.wikimedia.org.

useCases:
  - name: AI Grounding And RAG
    description: Ground LLM responses in encyclopedic content via page summaries, full HTML, or Wikimedia Enterprise structured contents - the primary use case for the proliferating community MCP servers.
  - name: Voice Assistants And Smart Speakers
    description: Siri, Alexa, and Google Assistant resolve open-domain questions through Wikipedia page summaries and Wikidata.
  - name: Encyclopedia Mirrors And Offline Reading
    description: Kiwix and the Wikipedia Apps consume the REST API plus periodic dumps to ship offline-readable copies to schools and constrained-connectivity regions.
  - name: Knowledge Graph Research
    description: Researchers query the Wikidata SPARQL endpoint to study entity relationships, citation networks, and cross-language alignment.
  - name: Bot Workflows (Anti-Vandalism, Citation, Categorization)
    description: Tens of thousands of community bots use the Action API to revert vandalism, fix references, and maintain categories.
  - name: News Aggregation
    description: Realtime EventStream of edits feeds breaking-news detection - the In-The-News template on Wikipedia main page is itself a data product.
  - name: Linked Data Publishing
    description: Wikidata is the open-license source-of-truth for identifiers (ISBN, ORCID, GeoNames, MusicBrainz) used across linked-data publishers.

integrations:
  - name: Wikidata
    description: Structured-data sister project; every Wikipedia page links to a Wikidata QID consumable via the SPARQL endpoint.
  - name: Wikimedia Commons
    description: Shared media repository accessed via the same Action API at commons.wikimedia.org/w/api.php.
  - name: schema.org
    description: Page summaries align with schema.org/Article; Wikidata properties map to schema.org types.
  - name: DBpedia
    description: Community knowledge graph built by extracting structured data from Wikipedia infoboxes.
  - name: Google Knowledge Graph
    description: Major downstream consumer of Wikipedia + Wikidata content for search result panels.
  - name: ChatGPT / Claude / Gemini
    description: All major LLMs were pre-trained on Wikipedia and retrieve via REST API or community MCPs at inference time.
  - name: VisualEditor / Parsoid
    description: WYSIWYG editor and the parser that powers the Core REST transforms.

solutions:
  - name: Open-Knowledge Foundation
    description: A non-profit, donor-funded resource serving 1.5B+ unique devices/month while keeping every byte free to redistribute under CC BY-SA / CC0.
  - name: Commercial High-Volume Tier
    description: Wikimedia Enterprise (Wikimedia LLC) offers SLA-backed Snapshot, On-demand, and Realtime APIs - the same content packaged for AI labs, search engines, and content platforms that need predictable throughput and update cadence.

maintainers:
  - FN: Kin Lane
    email: [email protected]