extractor.sh

Developer documentation

Search. Extract.
Done.

Search the public web or extract a known page through cacheable GET endpoints. Receive readable Markdown or normalized JSON.

curl --get 'https://extractor.sh/api/extract' \
  --data-urlencode 'url=https://example.com/article' \
  --data-urlencode 'format=markdown'

Start here

QuickstartMake your first Markdown and JSON requests.API referenceParameters, responses, fields, and errors.Web searchDiscover relevant pages with one GET request.News searchGet current coverage as a normalized article feed.Image searchFind openly licensed images with source and license metadata.Place searchResolve named places and addresses to normalized coordinates.FinanceSearch stock symbols and get native or converted market history.Hosted MCP serverConnect agents directly to search and extraction tools.Supported sourcesThe public page URLs you can submit.JSON schemaTypes, shared fields, product prices, and collection items.Limits and cachingRate limits, cache lifetimes, and safe usage.LimitationsCAPTCHAs, access controls, regions, and source boundaries.

Base URL

https://extractor.sh

No account is required for 10 successful uncached operations per IP each UTC day. New accounts receive a one-time welcome bonus of 1,000 credits. Signed-in website tools use the account balance automatically, while external integrations use the same balance with a Bearer key created in the dashboard. Do not put credentials, cookies, private URLs, or private data in query parameters.

Accounts and prepaid usage

Read the API-key guide for key creation and revocation, and the billing guide for exact pay-as-you-go top-ups and credit reversals.