Jina AI MCP Tools

A Model Context Protocol (MCP) server that integrates with Jina AI Search Foundation APIs.

Features

This MCP server provides access to the following Jina AI APIs:

• Web Reader - Extract content from web pages using r.jina.ai
• Web Search - Search the web using s.jina.ai or svip.jina.ai (configurable via --search-endpoint)

Prerequisites

1. Jina AI API Key (Optional) - Get a free API key from https://jina.ai/?sui=apikey for enhanced features
2. Node.js - Version 20 or higher

MCP Server

Using stdio Transport (Default)

For local integrations spawned by another process (e.g., Claude Desktop, VS Code, Cursor):

{
  "mcpServers": {
    "jina-mcp-tools": {
      "command": "npx",
      "args": [
        "jina-mcp-tools",
        "--transport", "stdio",
        "--tokens-per-page", "15000",
        "--search-endpoint", "standard"
      ],
      "env": {
        "JINA_API_KEY": "your_jina_api_key_here_optional"
      }
    }
  }
}

Using HTTP Transport

For remote server deployments accessible via HTTP:

Start the server:

# With API key
JINA_API_KEY=your_api_key npx jina-mcp-tools --transport http --host 127.0.0.1 --port 3000

# Without API key (reader tool only, binds to 127.0.0.1 by default)
npx jina-mcp-tools --transport http --port 3000

Connect from MCP clients:

• MCP Inspector: npx @modelcontextprotocol/inspector → http://localhost:3000/mcp
• Claude Code: claude mcp add --transport http jina-tools http://localhost:3000/mcp
• VS Code: code --add-mcp '{"name":"jina-tools","type":"http","url":"http://localhost:3000/mcp"}'

CLI Options:

• --transport - Transport type: stdio or http (default: stdio)
• --host - Host/interface to bind in HTTP mode (default: 127.0.0.1)
• --port - HTTP server port (default: 3000, only for HTTP transport)
• --tokens-per-page - Tokens per page for pagination (default: 15000)
• --search-endpoint - Search endpoint to use: standard (s.jina.ai) or vip (svip.jina.ai) (default: standard)
• --cache-size - Reader cache size in URLs (default: 50)
• --help - Show the built-in CLI help

HTTP Security

HTTP mode binds to 127.0.0.1 by default. For remote deployments, put the server behind TLS and set a bearer token:

JINA_MCP_HTTP_AUTH_TOKEN=change-me npx jina-mcp-tools --transport http --host 0.0.0.0 --port 3000

Clients must then send:

Authorization: Bearer change-me

Browser-origin requests are limited to localhost by default. Set JINA_MCP_ALLOWED_ORIGINS to a comma-separated allowlist for browser-based remote clients.

Proxy Environment Variables

To route outbound requests through a proxy, enable proxy environment support and provide the proxy URLs:

NODE_USE_ENV_PROXY=1 \
HTTP_PROXY=http://127.0.0.1:7890 \
HTTPS_PROXY=http://127.0.0.1:7890 \
NO_PROXY=localhost,127.0.0.1,::1 \
npx jina-mcp-tools --transport stdio

For MCP client configs, include the same environment variables:

{
  "mcpServers": {
    "jina-mcp-tools": {
      "command": "npx",
      "args": ["jina-mcp-tools"],
      "env": {
        "JINA_API_KEY": "your_jina_api_key_here_optional",
        "NODE_USE_ENV_PROXY": "1",
        "HTTP_PROXY": "http://127.0.0.1:7890",
        "HTTPS_PROXY": "http://127.0.0.1:7890",
        "NO_PROXY": "localhost,127.0.0.1,::1"
      }
    }
  }
}

Alternatively, start Node with NODE_OPTIONS=--use-env-proxy. Proxy URLs are only used when proxy environment support is enabled.

Available Tools

jina_reader

Extract and read web page content.

Parameters:

• url - URL to read (required)
• page - Page number for paginated content (default: 1)
• customTimeout - Timeout override in seconds (optional)

Features:

• Automatic pagination for large documents
• LRU cache (50 URLs) for instant subsequent page requests
• GitHub file URLs automatically converted to raw content URLs
• Tries Accept: text/markdown first for a maintained allowlist of markdown-capable docs/blog hosts, then falls back to r.jina.ai on fetch failure or empty responses

jina_search / jina_search_vip

Search the web. Returns partial content; use jina_reader for full content. Requires API key.

Tool registered depends on --search-endpoint:

• jina_search → standard (s.jina.ai, default)
• jina_search_vip → vip (svip.jina.ai)

Parameters:

• query - Search query (required)
• count - Number of results (default: 5)
• siteFilter - Limit to specific domain (e.g., "github.com")

License

MIT

Links

• GitHub: https://github.com/PsychArch/jina-mcp-tools
• Issues: https://github.com/PsychArch/jina-mcp-tools/issues