Build your own Qdrant vector store MCP server

by Jimleuk ✓ · 3,226 views · created 2025-04-21 · Engineering, AI RAG

What this workflow does

This AI workflow has 37 nodes and runs manually, when you click “Execute workflow” (plus 2 other triggers). Along the way it connects Embeddings OpenAI and Qdrant Vector Store. It also calls qdrant:6333 and api.openai.com over HTTP.

How it runs

Trigger: Manually, when you click “Execute workflow” (plus 2 other triggers).

1. Create Collection (HTTP Request): calls qdrant:6333 (PUT)
2. Operation (Switch): routes items into 5 branches
3. Create Facet Index (HTTP Request): calls qdrant:6333 (PUT)
4. List by Facet API (HTTP Request): calls qdrant:6333 (POST)
5. Insert Reviews (Qdrant Vector Store): stores documents in the Qdrant vector store, with embeddings: Embeddings OpenAI; document loader: Default Data Loader
6. Search Reviews (Qdrant Vector Store): searches the Qdrant vector store, with embeddings: Embeddings OpenAI
7. Get Embeddings (HTTP Request): calls api.openai.com (POST)
8. Preferences to Items (Code): runs custom JavaScript code
9. Get Insert Response (Edit Fields (Set)): sets 1 field: response
10. Get Search Response (Aggregate): combines items into one

…plus 15 more steps. See the full node list below.

Services it uses

Embeddings OpenAIQdrant Vector Store

HTTP calls to: qdrant:6333, api.openai.com

Before you upgrade to n8n 3.0

Nodes used (16)

• Call n8n Workflow Tool@n8n/n8n-nodes-langchain.toolWorkflow
• MCP Server Trigger@n8n/n8n-nodes-langchain.mcpTrigger
• Execute Workflow Triggern8n-nodes-base.executeWorkflowTrigger
• Switchn8n-nodes-base.switch
• HTTP Requestn8n-nodes-base.httpRequest
• Coden8n-nodes-base.code
• Aggregaten8n-nodes-base.aggregate
• Ifn8n-nodes-base.if
• Edit Fields (Set)n8n-nodes-base.set
• Embeddings OpenAI@n8n/n8n-nodes-langchain.embeddingsOpenAi
• Default Data Loader@n8n/n8n-nodes-langchain.documentDefaultDataLoader
• Recursive Character Text Splitter@n8n/n8n-nodes-langchain.textSplitterRecursiveCharacterTextSplitter
• Qdrant Vector Store@n8n/n8n-nodes-langchain.vectorStoreQdrant
• Split Outn8n-nodes-base.splitOut
• Filtern8n-nodes-base.filter
• Manual Triggern8n-nodes-base.manualTrigger

From the author

This n8n demonstrates how to build your own Qdrant MCP server to extend its functionality beyond that of the official implementation. This n8n implementation exposes other cool API features from Qdrant such as facet search, grouped search and recommendations…

View the original on n8n.io → Check your own copy