Skip to content

Reading pages ​

POST /v1/extract returns the main text of up to 10 web pages, ready to pass to a model. Use it after a search when a snippet isn't enough.

bash
curl https://api.usefully.info/v1/extract \
  -H "Authorization: Bearer $USEFULLY_API_KEY" -H "Content-Type: application/json" \
  -d '{"urls": ["https://example.com/blog/post"], "max_chars": 20000}'
FieldNotes
urls1–10 http(s) URLs
max_charsTruncate each page (default 100,000)

Each result has a status of ok or failed:

  • ok results have a page with title, content, final_url and, when known, published_at.
  • failed results have an error. Failed pages aren't billed.

Pages cost $1 per 1,000. Keys need the search scope.

If you only need page text for search results, pass "include_content": true to /v1/search and skip the second call.