- Search the web, news and images.
- Scrape any public page as markdown, HTML, links, a screenshot or JSON.
- Map a site’s pages, crawl a site, or batch scrape many pages.
- Extract the fields you name from a page.
- Read a site’s own data, like a Reddit search, a YouTube transcript or Google Maps places, through one of 103 parsers.
Connect your agent
https://mcp.stophy.dev/mcp instead and sign in with your browser. MCP server covers Cursor, Claude Desktop and other clients.
Next
Quickstart
Make a call from cURL, TypeScript or Python.
Parsers
Read a site’s own data as JSON.
Pricing
Most calls cost 1 credit, and you start with 5,000 free.
API reference
Every endpoint, with its inputs, outputs and price.