markdown returns the text of a page as Markdown. Headings, lists, tables, links and code blocks keep their shape, so you can pass the result to a model or a search index as it is. It is the default format: a request without formats returns markdown.
Make a request
Get an API key in the Quickstart. This request reads a page as Markdown.Read the response
The Markdown is indata.markdown. This real response is cut after the first paragraph. metadata describes the page. See Page metadata.
linkStyle), and to cut a long page or read it part by part (maxChars and offset).
Options
Every other field that shapes the Markdown is a top-level field of the request and is described in Content filtering, Waiting and rendering and Location and headers.
Limits and errors
- A PDF always comes back as Markdown. See PDFs and documents. A CSV or an Excel
.xlsxfile comes back as Markdown tables. - A YouTube link returns the video’s title, channel, date, length, views, likes and description. See YouTube.
- A page that answers with a 404 or 410 status, or cannot be read, costs nothing.
Price
$1 per 1,000 pages.markdown adds nothing to the price of the page.
Related
- HTML returns the page as HTML instead.
- Highlights returns only the passages that match a query.
- Content filtering keeps or removes parts of the page.
- Scrape overview lists every format.