Web Research

The Web MCP server enables retrieving and searching web content with Jina AI.

Main Features

Reading Websites

• read_url: Reads and processes website content
  • Converts HTML to structured markdown
  • Removes ads and navigation
  • Extracts the main content
  • Preserves important formatting

Web Search

• web_search: Searches the internet for information
  • Relevant search results
  • Optional: Complete content of found pages
  • Current information from the web

Use Cases

Extract Content

"Read the article on example.com/blog"
"What's on XYZ's product page?"
"Summarize the content of this news page"

Information Search

"Search for the latest AI developments"
"Find information about quantum computers"
"What are the current stock prices?"

Research

"Compare prices from different vendors"
"Collect reviews for this product"
"Find technical documentation for this API"

How It Works

Jina Reader

• Intelligent content extraction
• Removes disruptive elements (ads, navigation)
• Maintains structure (headings, lists, tables)
• Markdown output for better readability

Jina Search

• High-quality search results
• Optional content extraction
• Up-to-date information

Benefits

• Clean, structured content
• No browser automation needed
• Fast processing
• Reliable extraction

Configuration

Requires a Jina API key as environment variable JINA_API_KEY.