---
title: "Qdrant MCP Server for ToolPiper · ModelPiper"
description: "Store and search vector embeddings as a semantic memory. Qdrant is a vector database; this server turns it into a semantic memory the assistant can write to and search. Store snippets of text with metadata, then retrieve"
canonical: "https://modelpiper.com/mcp-tools/qdrant"
---

# Qdrant MCP Server for ToolPiper · ModelPiper

> Store and search vector embeddings as a semantic memory. Qdrant is a vector database; this server turns it into a semantic memory the assistant can write to and search. Store snippets of text with metadata, then retrieve

Q

# Qdrant MCP Server

✓ verified

Store and search vector embeddings as a semantic memory.

Install Qdrant in ToolPiper

Don't have ToolPiper? We'll take you to the download.

[Open repository →](https://github.com/qdrant/mcp-server-qdrant)

Publisher

[Qdrant](https://github.com/qdrant/mcp-server-qdrant)

License

Apache-2.0

Transport

Local subprocess

Tools

2

## What this server does

Qdrant is a vector database; this server turns it into a semantic memory the assistant can write to and search. Store snippets of text with metadata, then retrieve the most relevant ones by meaning rather than keyword. It is the building block for long-term recall across conversations: save decisions, facts, or code patterns once, and the model finds them later by similarity. Works against a local Qdrant or a Qdrant Cloud cluster. The server handles embedding generation with a sensible default model.

## Tools included

-   qdrant\_store
    
    Store a piece of text with its embedding and metadata.
    
-   qdrant\_find
    
    Find the most relevant stored entries for a query.
    

The authoritative list comes from the server's tools/list at runtime. This catalog entry may lag the upstream until the next ToolPiper release.

## Install in ToolPiper

ToolPiper handles the JSON configuration. Click Install, fill in any required tokens, and the server is ready in your chat.

Install Qdrant in ToolPiper

Don't have ToolPiper? We'll take you to the download.

## Install in another client

Install in another MCP client (Claude Desktop, Cursor, Continue, Cline)

Claude Desktop Cursor Continue Cline

Add this to ~/Library/Application Support/Claude/claude\_desktop\_config.json.

```
{
  "mcpServers": {
    "qdrant": {
      "command": "uvx",
      "args": [
        "mcp-server-qdrant"
      ]
    }
  }
}
```

Copy

## Example prompts

-   "Store this architecture decision so we can recall it later."
    
    Copy
-   "Find anything we have saved about how the auth service handles token refresh."
    
    Copy
-   "What did we decide about the database migration strategy?"
    
    Copy

## Security notes

For Qdrant Cloud the API key gates access to your cluster; treat it as a credential. The collection holds whatever text the assistant stores, in plaintext, so be deliberate about what gets written there, the same way you would with any memory store. A local instance needs no key but is only as secure as the machine it runs on.

## Related tools

-   [
    
    P
    
    ### Postgres
    
    ✓ verified
    
    Query Postgres databases with read-only SQL.
    
    DataLocal subprocess1 tool
    
    
    
    
    
    ](/mcp-tools/postgres)
-   [
    
    S
    
    ### SQLite
    
    ✓ verified
    
    Query SQLite databases and run business-intelligence prompts.
    
    DataLocal subprocess4 tools
    
    
    
    
    
    ](/mcp-tools/sqlite)
-   [
    
    R
    
    ### Redis
    
    ✓ verified
    
    Interact with Redis key-value stores from the assistant.
    
    DataLocal subprocess4 tools
    
    
    
    
    
    ](/mcp-tools/redis)
-   [
    
    A
    
    ### AWS Knowledge Base
    
    ✓ verified
    
    Retrieve answers from an AWS Bedrock Knowledge Base.
    
    DataLocal subprocess1 tool
    
    
    
    
    
    ](/mcp-tools/aws-kb-retrieval)

[← Back to all MCP tools](/mcp-tools)