Open-source vector database for similarity search and AI.
## Milvus MCP Server: Vector Database for AI Applications The **Milvus MCP Server** integrates the open-source vector database Milvus into Google Antigravity. This high-performance database is designed for similarity search and AI applications, supporting billions of vectors with millisecond query latency. ### Why Milvus MCP? Milvus powers production AI applications: - **Massive Scale**: Billions of vectors with fast search - **Multiple Indexes**: IVF, HNSW, DiskANN, and more - **Hybrid Search**: Combine vector and scalar filtering - **Cloud Native**: Kubernetes-native architecture - **Antigravity Native**: AI-assisted vector operations ### Key Features #### 1. Vector Operations ```python from pymilvus import connections, Collection, FieldSchema, CollectionSchema, DataType # Connect to Milvus connections.connect("default", host="localhost", port="19530") # Define collection schema fields = [ FieldSchema(name="id", dtype=DataType.INT64, is_primary=True), FieldSchema(name="embedding", dtype=DataType.FLOAT_VECTOR, dim=768), FieldSchema(name="text", dtype=DataType.VARCHAR, max_length=1000) ] schema = CollectionSchema(fields, description="Document embeddings") collection = Collection("documents", schema) # Insert vectors collection.insert([ids, embeddings, texts]) ``` #### 2. Similarity Search ```python # Create index for fast search index_params = { "metric_type": "COSINE", "index_type": "HNSW", "params": {"M": 16, "efConstruction": 256} } collection.create_index("embedding", index_params) collection.load() # Search similar vectors results = collection.search( data=[query_embedding], anns_field="embedding", param={"metric_type": "COSINE", "params": {"ef": 64}}, limit=10, output_fields=["text"] ) ``` #### 3. Hybrid Filtering ```python # Combine vector search with filters results = collection.search( data=[query_embedding], anns_field="embedding", param={"ef": 64}, limit=10, expr="category == 'technology' and date > '2024-01-01'", output_fields=["text", "category", "date"] ) ``` ### Configuration ```json { "mcpServers": { "milvus": { "command": "npx", "args": ["-y", "@anthropic/mcp-milvus"], "env": { "MILVUS_HOST": "localhost", "MILVUS_PORT": "19530", "MILVUS_TOKEN": "your-token" } } } } ``` ### Use Cases **Semantic Search**: Build search systems that understand meaning, not just keywords. **Recommendation Systems**: Find similar items based on embedding similarity for personalized recommendations. **Image/Video Search**: Search media by content similarity using vision model embeddings. The Milvus MCP Server brings production-grade vector search to Antigravity applications.
{
"mcpServers": {
"milvus": {}
}
}