Extract API Quickstart
Fetch any webpage or PDF and parse it into clean Markdown or HTML.
The Extract API fetches any webpage or PDF and parses it into clean Markdown or HTML. Send one URL per request and read the page from contents.full_page.
Get an API key at you.com/platform. The samples below read YDC_API_KEY. See the site Quickstart for the full API tour and API key management for best practices.
Extract a Webpage
The response is one object:
Extract a PDF
Send a PDF URL to the same endpoint. Tables come back with their rows and columns intact. Use parse_effort: "lite" for text-based PDFs to skip the per-page charge, or keep the default standard when charts and images carry the information you need. Raise timeout (up to 100 seconds) for long documents.
If you have a query rather than a URL, use full page extraction on the Web Search API instead.