ncasfl/Qdrant-MCP
CP server exposing Qdrant vector DB store and GPU-accelerated embedding pipeline as tools for Claude Code, Claude.ai, and OpenClaw.
Platform-specific configuration:
{
"mcpServers": {
"Qdrant-MCP": {
"command": "npx",
"args": [
"-y",
"Qdrant-MCP"
]
}
}
}Add the config above to .claude/settings.json under the mcpServers key.
<p align="center"> <h1 align="center">Qdrant MCP Server</h1> <p align="center"> Semantic search, document ingestion, and collection management over <a href="https://qdrant.tech">Qdrant</a> — powered by <a href="https://modelcontextprotocol.io">MCP</a> </p> </p>
<p align="center"> <a href="#tools">Tools</a> • <a href="#quickstart">Quickstart</a> • <a href="#embedding-providers">Embedding Providers</a> • <a href="#client-configuration">Client Configuration</a> • <a href="#configuration-reference">Configuration</a> • <a href="#docker">Docker</a> • <a href="#troubleshooting">Troubleshooting</a> </p>
---
An MCP server that exposes your Qdrant vector database as tools for AI assistants. Ingest documents, search semantically, and manage collections — all through natural conversation.
Built with FastMCP (Python). Supports any OpenAI-compatible embedding provider — local (Ollama, llama.cpp) or cloud (OpenAI, Anthropic, Google, Azure).
Once the server is running and connected to your MCP client, interact through natural conversation:
> You: Search my documents for contract termination clauses
Search results for: contract termination clauses
Collection: documents | Results: 3
[1] 0.8234 — services-agreement.pdf p.12 (chunk 23)
Either party may terminate this Agreement upon thirty (30) days written
noLoading reviews...