Vector databases

Learn more about Mintlify

Enter your email to receive updates about new features and product releases.

Vector databases store embeddings and enable similarity search.

PgVector (PostgreSQL)

Recommended for production:

from agno.vectordb.pgvector import PgVector

vector_db = PgVector(
    table_name="knowledge",
    db_url="postgresql://localhost/agno"
)

Good for development:

from agno.vectordb.chromadb import ChromaDb

vector_db = ChromaDb(
    collection="docs",
    path="./chroma_db"
)

from agno.vectordb.lancedb import LanceDb

vector_db = LanceDb(
    table_name="knowledge",
    uri="./lancedb"
)

Supported: Qdrant, Pinecone, Weaviate, Milvus, and more. See storage documentation for setup.

⌘I