Jina AI

Provides semantic, image, and cross-modal search capabilities via MCP for Jina AI neural search.

javascript

4

GitHub Stars

javascript

Language

6 months ago

First Indexed

2 months ago

Catalog Refreshed

Documentation & install

Readme and setup notes from the catalogue, plus a client-ready config you can copy for your MCP host.

Installation

Add the following to your MCP client configuration file.

Configuration

View docs

{
  "mcpServers": {
    "sheshiyer-jina-ai-mcp-multimodal-search": {
      "command": "node",
      "args": [
        "/path/to/jina-ai-mcp/build/index.js"
      ],
      "env": {
        "JINA_API_KEY": "your_api_key_here"
      }
    }
  }
}

You deploy a JavaScript MCP server that integrates Jina AI’s neural search capabilities, enabling semantic, image, and cross-modal searches through a simple interface. This server exposes ready-to-run tooling so you can query text, images, or both, without building the underlying search models yourself.

How to use

You interact with the Jina AI MCP Server using an MCP client to perform semantic, image, and cross-modal searches. You’ll start by ensuring your API key is configured, then call the provided tools to search across your collections. Use the semantic_search tool for natural language queries on text, image_search for visually similar images via URL, and cross_modal_search to combine text and image modalities.

Typical usage patterns you can achieve:

Find documents related to a natural language query using semantic_search
Locate visually similar images by providing an image URL through image_search
Perform cross-modal queries that translate text into images or images into text via cross_modal_search

How to install

prerequisites: Node.js 16 or higher and a Jina AI API key.

Clone the project and navigate into the MCP directory.

git clone <repository-url>
cd jina-ai-mcp

Install dependencies.

npm install

Create a configuration file with your API key.

JINA_API_KEY=your_api_key_here

Build the server so it’s ready to run.

npm run build

Configuration

Configure the MCP server so your client can start and reach the Jina AI backend. The following is the provided MCP configuration that runs the server locally via Node.js.

{
  "mcpServers": {
    "jina-ai": {
      "command": "node",
      "args": [
        "/path/to/jina-ai-mcp/build/index.js"
      ],
      "env": {
        "JINA_API_KEY": "your_api_key_here"
      }
    }
  }
}

Security and environment

Keep your Jina API key secure. Do not hard-code the key in public repositories. Use environment variables or secret management to supply the API key to the MCP server in production.

Troubleshooting and notes

If you encounter invalid API key errors, verify that the JINA_API_KEY you configured matches the key issued by your Jina AI account. Check for missing parameters or network issues if you see rate limit or connectivity errors. Ensure the build completed successfully before starting the MCP server.

Tools and endpoints overview

The server exposes three primary tools you can call through your MCP client:

Semantic Search: query text to retrieve semantically relevant documents from a collection.
Image Search: supply an image URL to find visually similar images within a collection.
Cross-Modal Search: perform text-to-image or image-to-text searches across your data.

Available tools

semantic_search

Perform semantic/neural search on text documents by submitting a natural language query and receiving ranked results.

image_search

Search for visually similar images by providing an image URL and optional collection and limit parameters.

cross_modal_search

Run cross-modal searches to translate between text and images, supporting text2image and image2text modes.