MCP server for Qdrant vector database
## Qdrant MCP Server: Open Source Vector Database The **Qdrant MCP Server** integrates the high-performance Qdrant vector database into Google Antigravity. This open-source solution offers advanced filtering, payload storage, and blazing-fast similarity search for production AI applications. ### Why Qdrant MCP? Qdrant combines performance with flexibility: - **Open Source**: Self-host or use cloud - **Rich Filtering**: Complex payload queries - **Quantization**: Reduce memory usage 4x - **Multivector**: Multiple vectors per point - **HNSW Index**: Fast approximate search ### Key Features #### 1. Collection Management ```python from qdrant_client import QdrantClient from qdrant_client.models import Distance, VectorParams client = QdrantClient(url="http://localhost:6333") # Create collection client.create_collection( collection_name="documents", vectors_config=VectorParams(size=768, distance=Distance.COSINE) ) ``` #### 2. Upsert and Search ```python from qdrant_client.models import PointStruct # Add points with payload client.upsert( collection_name="documents", points=[ PointStruct( id=1, vector=embedding1, payload={"title": "ML Guide", "category": "tech", "date": "2024-01-15"} ) ] ) # Search with filters results = client.search( collection_name="documents", query_vector=query_embedding, query_filter=Filter( must=[FieldCondition(key="category", match=MatchValue(value="tech"))] ), limit=10 ) ``` #### 3. Advanced Filtering ```python from qdrant_client.models import Filter, FieldCondition, Range # Complex filter queries results = client.search( collection_name="documents", query_vector=query_embedding, query_filter=Filter( must=[ FieldCondition(key="category", match=MatchValue(value="tech")), FieldCondition(key="date", range=Range(gte="2024-01-01")) ] ), limit=10 ) ``` ### Configuration ```json { "mcpServers": { "qdrant": { "command": "npx", "args": ["-y", "@anthropic/mcp-qdrant"], "env": { "QDRANT_URL": "http://localhost:6333", "QDRANT_API_KEY": "your-api-key" } } } } ``` ### Use Cases **Filtered Search**: Combine semantic similarity with rich metadata filtering. **Recommendation Systems**: Build personalized recommendations with payload context. **Multi-Tenant RAG**: Isolate data with collection-level or payload-based separation. The Qdrant MCP Server brings flexible vector search to Antigravity.
{
"mcpServers": {
"qdrant": {
"mcpServers": {
"qdrant": {
"env": {
"QDRANT_URL": "http://localhost:6333",
"QDRANT_API_KEY": "YOUR_QDRANT_API_KEY"
},
"args": [
"qdrant-mcp"
],
"command": "uvx"
}
}
}
}
}