Jina AI MCP

How to Install and Use Jina AI MCP

Jina AI MCP is a remote server that offers access to powerful tools like Reader, Embeddings, and Reranker APIs. These tools help you work with web pages, search the internet, and analyze documents. Here’s a simple guide to help you install and use Jina AI MCP quickly.

Setting Up the Jina AI MCP Server

To get started, you need to tell your client where to find the Jina AI MCP server. This is done by adding a configuration with the server URL and an optional API key.

First, here’s the basic setup if your client supports remote MCP servers:

{
  "mcpServers": {
    "jina-mcp-server": {
      "url": "https://mcp.jina.ai/v1",
      "headers": {
        "Authorization": "Bearer ${JINA_API_KEY}"  // optional
      }
    }
  }
}

Replace ${JINA_API_KEY} with your actual API key, which you can get for free from https://jina.ai.

If your client does not support remote MCP servers yet, you need to use the mcp-remote local proxy. Here is the configuration example:

{
  "mcpServers": {
    "jina-mcp-server": {
      "command": "npx",
      "args": [
        "mcp-remote",
        "https://mcp.jina.ai/v1",
        "--header",
        "Authorization: Bearer ${JINA_API_KEY}"
      ]
    }
  }
}

This setup lets you connect to the remote MCP server through a local proxy that speaks the required protocol.

Using Jina AI MCP with Claude Code and OpenAI Codex

If you work with Claude Code, here’s how to add the MCP server:

claude mcp add -s user --transport http jina https://mcp.jina.ai/v1 \
  --header "Authorization: Bearer ${JINA_API_KEY}"

Make sure to remove any old server using the /sse transport before adding this one.

For OpenAI Codex users, you can add this block to your configuration file ~/.codex/config.toml:

[mcp_servers.jina-mcp-server]
command = "npx"
args = [
  "-y",
  "mcp-remote",
  "https://mcp.jina.ai/v1",
  "--header",
  "Authorization: Bearer ${JINA_API_KEY}"
]

This makes Codex connect via the proxy to the Jina AI MCP server.

Filtering Tools to Save Context Space

Jina AI MCP offers many tools, but some clients have limited context space for tool names and descriptions. You can choose only the tools you want by adding filters to the server URL. This helps keep your client fast and efficient.

For example, to exclude parallel tools, use this configuration:

{
  "mcpServers": {
    "jina-mcp-server": {
      "url": "https://mcp.jina.ai/v1?exclude_tags=parallel",
      "headers": {
        "Authorization": "Bearer ${JINA_API_KEY}"
      }
    }
  }
}

Or to include only search and reading tools:

{
  "mcpServers": {
    "jina-mcp-server": {
      "url": "https://mcp.jina.ai/v1?include_tags=search,read",
      "headers": {
        "Authorization": "Bearer ${JINA_API_KEY}"
      }
    }
  }
}

You can also exclude specific tools with exclude_tools like this:

{
  "mcpServers": {
    "jina-mcp-server": {
      "url": "https://mcp.jina.ai/v1?exclude_tools=search_images,deduplicate_images",
      "headers": {
        "Authorization": "Bearer ${JINA_API_KEY}"
      }
    }
  }
}

Troubleshooting Common Issues

If you face troubles like stuck tool calling loops, red dots on MCP status, or missing tools, here are some quick tips:

• Increase your model’s context size to avoid loops.
• Remove and re-add the MCP server to get fresh tools.
• For Windows, fix spacing issues in arguments by moving them to environment variables.

With these, you can keep your Jina AI MCP running smoothly.

---

These steps show you how to install and start using the Jina AI MCP server easily. Just add the right config, use or filter tools as needed, and explore powerful search and reading capabilities right away.