Qdrant MCP Server — Hosted Api Connectors, Any AI Agent

mcp-qdrant — Qdrant MCP Server

Search, upsert, and manage vector collections in your Qdrant database from your AI agents.

Qdrant is the open-source vector search engine built for AI applications — powering semantic search, RAG pipelines, and recommendation systems at scale. This MCP server gives your AI agents direct access to your Qdrant instance: creating collections, upserting vectors with payloads, running similarity searches, and managing points — all through natural language.

Live endpoint: https://mcp.aerostack.dev/s/aerostack/mcp-qdrant

---

What You Can Do

• Create and manage vector collections with configurable dimensions and distance metrics
• Upsert points with vectors and metadata payloads for semantic indexing
• Run similarity searches against your vectors to power RAG retrieval and recommendations
• Scroll, count, and filter points to explore and audit your vector data

Available Tools

Tool	Description
list_collections	List all collections in the database
get_collection	Get collection info including vector count, config, and status
create_collection	Create a new collection with vector size and distance metric
delete_collection	Delete a collection and all its data
upsert_points	Upsert points (vectors + payload) into a collection
search	Search for similar vectors with filters and score thresholds
get_points	Get points by their IDs
delete_points	Delete points by IDs or filter
scroll	Scroll through points with optional filter and pagination
count	Count points in a collection with optional filter

Configuration

Variable	Required	Description	How to Get
QDRANT_URL	Yes	Your Qdrant instance URL (e.g. https://xyz.us-east4-0.gcp.cloud.qdrant.io:6333)	cloud.qdrant.io → Your Cluster → copy Cluster URL
QDRANT_API_KEY	Yes	Qdrant API key for authentication	cloud.qdrant.io → Your Cluster → Data Access Control → create or copy API Key

Quick Start

Add to Aerostack Workspace

1. Go to aerostack.dev → Your Project → MCPs
2. Search for "Qdrant" and click Add to Workspace
3. Add QDRANT_URL and QDRANT_API_KEY under Project → Secrets

Once added, every AI agent in your workspace can call Qdrant tools automatically — no per-user setup needed.

Example Prompts

"List all collections in my Qdrant database"
"Create a new collection called 'documents' with 1536-dimensional vectors using cosine distance"
"Search the 'documents' collection for vectors similar to this embedding, return top 5 results"
"Count how many points are in the 'products' collection where category is 'electronics'"

Direct API Call

curl -X POST https://mcp.aerostack.dev/s/aerostack/mcp-qdrant \
  -H 'Content-Type: application/json' \
  -H 'X-Mcp-Secret-QDRANT-URL: https://your-cluster.cloud.qdrant.io:6333' \
  -H 'X-Mcp-Secret-QDRANT-API-KEY: your-api-key' \
  -d '{"jsonrpc":"2.0","id":1,"method":"tools/call","params":{"name":"list_collections","arguments":{}}}'

License

MIT