Qdrant With OpenAI Embeddings

🚀 MCP Qdrant Server with OpenAI Embeddings

This MCP server leverages the Qdrant vector database and OpenAI embeddings to provide vector search capabilities.

🚀 Quick Start

This MCP server uses the Qdrant vector database and OpenAI embeddings to offer vector search functionality.

✨ Features

• Conduct semantic searches in Qdrant collections using OpenAI embeddings.
• List all available Qdrant collections.
• View information about a specific Qdrant collection.

📦 Installation

Prerequisites

• Python 3.10+ installed.
• A Qdrant instance (either local or remote).
• An OpenAI API key.

Steps

1. Clone this repository:

git clone https://github.com/yourusername/mcp-qdrant-openai.git
cd mcp-qdrant-openai

2. Install the dependencies:

pip install -r requirements.txt

📚 Documentation

Configuration

Set the following environment variables:

• OPENAI_API_KEY: Your OpenAI API key.
• QDRANT_URL: The URL of your Qdrant instance (default: "http://localhost:6333").
• QDRANT_API_KEY: Your Qdrant API key (if applicable).

Usage

Run the server directly:

python mcp_qdrant_server.py

Run using the MCP CLI:

mcp dev mcp_qdrant_server.py

Install on Claude Desktop:

mcp install mcp_qdrant_server.py --name "Qdrant-OpenAI"

Available Tools

query_collection

Perform a semantic search on a Qdrant collection using OpenAI embeddings.

• collection_name: The name of the Qdrant collection to search.
• query_text: A natural language search query.
• limit: The maximum number of results to return (default: 5).
• model: The OpenAI embedding model to use (default: text-embedding-3-small).

list_collections

List all available Qdrant collections.

collection_info

Get information about a specific collection.

• collection_name: The name of the collection to get information about.

💻 Usage Examples

Example Usage on Claude Desktop

Once installed on Claude Desktop, you can use the tools like this:

What collections do I have in my Qdrant database?

Search for content about climate change in my "documents" collection.

Show information about the "articles" collection.