Turn a URL into readable Markdown
Give an AI agent page text without forwarding its cookies or credentials.
How it works
The reader accepts a public HTTPS URL and returns UTF-8 Markdown. It can also return JSON with the title, final URL, media type, and a truncation flag.
HTML, PDF, Markdown, plain text, JSON, CSV, XML, DOCX, XLSX, and PPTX are supported. Add render=true when the page needs an isolated browser to build its content.
Requests are bounded by size and time limits. Scanned PDFs need OCR, which this tool does not provide.
Quick start
Public API calls use HTTPS GET. This example uses curl to encode query values. Run creation and relay requests deliberately; they can change state.
curl --get 'https://pageplaintext.com/v1/read' \
--data-urlencode 'url=https://example.com' \
--data-urlencode 'format=json'Request parameters
| Parameter | What it does |
|---|---|
url | Required public page or document URL. |
format | markdown by default; json adds title, final URL, media type, and truncation metadata. |
render | Set true for pages that require browser rendering. |
Response
JSON includes markdown, title, url, source_content_type, and truncated. The default response is Markdown text.
Complete the flow
The default response is Markdown. For title, final URL, media type, and a truncation flag, repeat the same GET with format=json. No token is required.