Jina AI Search

You run this MCP server to expose Jina AI’s Search Foundation API through the Model Context Protocol, enabling AI assistants and applications to perform fast, scalable search, reading, and knowledge retrieval via MCP clients.

How to use

You connect to the MCP server from your MCP client (for example Claude Desktop) to perform searches, read web content, and explore knowledge sources like DeepWiki. Initiate the client configuration to point at the MCP server endpoints or local runtimes shown below, then issue queries or content requests as you work.

How to install

Prerequisites: Python 3.8 or later and a Python package manager (pip). Ensure you have a valid Jina API key to access the Jina AI services.

Step-by-step setup for local development and running the MCP server:

# 1. Create and activate a virtual environment
python -m venv venv
# On Windows:
venv\Scripts\activate
# On macOS/Linux:
source venv/bin/activate

# 2. Install dependencies
pip install -r requirements.txt

# 3. Start the MCP server locally
python -m jina_mcp.server

Configuration

Set your Jina API key in the environment to enable authenticated access to Jina’s services.

JINA_API_KEY=your_jina_api_key_here

MCP Server Setup

You can connect from an MCP client in two ways: a remote client configuration (via Claude Desktop) and a local runtime for development.

{
  "mcpServers": {
    "jina_desktop": {
      "command": "cmd",
      "args": [
        "/c",
        "npx",
        "-y",
        "supergateway",
        "--sse",
        "https://jina-mcp.onrender.com/sse",
        "--oauth2Bearer=your-jina-api-key"
      ]
    }
  }
}

{
  "mcpServers": {
    "jina_local": {
      "command": "python",
      "args": ["-m", "jina_mcp.server"],
      "cwd": "/path/to/jina-mcp",
      "env": {
        "JINA_API_KEY": "your-api-key-here"
      }
    }
  }
}

Usage Examples

Use simple, direct prompts to explore search, reading, and knowledge tasks. Examples include asking for latest developments in AI, analyzing web pages, or querying DeepWiki for explanations.

# Basic search prompt examples
"Search for the latest developments in artificial intelligence"
"Find information about sustainable energy solutions"
"What are the recent breakthroughs in quantum computing?"

# DeepWiki queries
"Ask deepwiki about machine learning"
"Explain neural networks with examples"
"What is Python programming?"

# Web content reading
"Read https://example.com/blog-post"
"Summarize the content from https://research-paper-url.com"

API Documentation

Access to Jina AI’s search and reader capabilities is provided through the MCP server, enabling you to perform search, reading, and knowledge tasks with MCP clients.

Development

To contribute or customize the MCP server, follow typical Python project workflows: install dependencies, run tests, and format code. Ensure your API key remains secure and do not commit it into version control.

Support

If you need help, reach out through the support channels provided by the project for assistance with MCP setup, authentication, or usage questions.

Available tools

search

Execute a search against Jina AI's search foundation to retrieve relevant results from multiple sources.

read

Read and extract content from web pages or documents to summarize or analyze information.

deepwiki

Query DeepWiki-enabled sources for AI-enhanced explanations and knowledge retrieval.

sse

Use Server-Sent Events for real-time streaming of MCP responses to your client.