SurrealDB Vector Database

Use SurrealDB as a vector database for your Knowledge Base.

Setup

docker run --rm \
  --pull always \
  -p 8000:8000 \
  surrealdb/surrealdb:latest \
  start \
  --user root \
  --pass root

or

./cookbook/scripts/run_surrealdb.sh

Example

agent_with_knowledge.py

from agno.agent import Agent
from agno.knowledge.embedder.openai import OpenAIEmbedder
from agno.knowledge.knowledge import Knowledge
from agno.vectordb.surrealdb import SurrealDb
from surrealdb import Surreal

# SurrealDB connection parameters
SURREALDB_URL = "ws://localhost:8000"
SURREALDB_USER = "root"
SURREALDB_PASSWORD = "root"
SURREALDB_NAMESPACE = "test"
SURREALDB_DATABASE = "test"

# Create a client
client = Surreal(url=SURREALDB_URL)
client.signin({"username": SURREALDB_USER, "password": SURREALDB_PASSWORD})
client.use(namespace=SURREALDB_NAMESPACE, database=SURREALDB_DATABASE)

surrealdb = SurrealDb(
    client=client,
    collection="recipes",  # Collection name for storing documents
    efc=150,  # HNSW construction time/accuracy trade-off
    m=12,  # HNSW max number of connections per element
    search_ef=40,  # HNSW search time/accuracy trade-off
)

def sync_demo():
    """Demonstrate synchronous usage of SurrealDb"""
    knowledge_base = Knowledge(
        vector_db=surrealdb,
        embedder=OpenAIEmbedder(),
    )

    # Load data synchronously
    knowledge_base.insert(
        url="https://agno-public.s3.amazonaws.com/recipes/ThaiRecipes.pdf"
    )

    # Create agent and query synchronously
    agent = Agent(knowledge=knowledge_base)
    agent.print_response(
        "What are the 3 categories of Thai SELECT is given to restaurants overseas?",
        markdown=True,
    )

if __name__ == "__main__":
    # Run synchronous demo
    print("Running synchronous demo...")
    sync_demo()

Async Support ⚡

SurrealDB also supports asynchronous operations, enabling concurrency and leading to better performance.

async_surrealdb_db.py

import asyncio

from agno.agent import Agent
from agno.knowledge.embedder.openai import OpenAIEmbedder
from agno.knowledge.knowledge import Knowledge
from agno.vectordb.surrealdb import SurrealDb
from surrealdb import AsyncSurreal

# SurrealDB connection parameters
SURREALDB_URL = "ws://localhost:8000"
SURREALDB_USER = "root"
SURREALDB_PASSWORD = "root"
SURREALDB_NAMESPACE = "test"
SURREALDB_DATABASE = "test"

# Create a client
client = AsyncSurreal(url=SURREALDB_URL)

surrealdb = SurrealDb(
async_client=client,
collection="recipes",  # Collection name for storing documents
efc=150,  # HNSW construction time/accuracy trade-off
m=12,  # HNSW max number of connections per element
search_ef=40,  # HNSW search time/accuracy trade-off
)

async def async_demo():
"""Demonstrate asynchronous usage of SurrealDb"""

await client.signin({"username": SURREALDB_USER, "password": SURREALDB_PASSWORD})
await client.use(namespace=SURREALDB_NAMESPACE, database=SURREALDB_DATABASE)

knowledge_base = Knowledge(
    vector_db=surrealdb,
    embedder=OpenAIEmbedder(),
)

await knowledge_base.ainsert(
    url="https://agno-public.s3.amazonaws.com/recipes/ThaiRecipes.pdf"
)

agent = Agent(knowledge=knowledge_base)
await agent.aprint_response(
    "What are the 3 categories of Thai SELECT is given to restaurants overseas?",
    markdown=True,
)

if __name__ == "__main__":
# Run asynchronous demo
print("\nRunning asynchronous demo...")
asyncio.run(async_demo())

Using aload() and aprint_response() with asyncio provides non-blocking operations, making your application more responsive under load.

SurrealDB Params

Parameter	Type	Default	Description
client	Optional[Union[BlockingWsSurrealConnection, BlockingHttpSurrealConnection]]	None	A blocking connection, either HTTP or WS
async_client	Optional[Union[AsyncWsSurrealConnection, AsyncHttpSurrealConnection]]	None	An async connection, either HTTP or WS
collection	str	"documents"	Collection name to store documents
distance	Distance	Distance.cosine	Distance metric to use (cosine, l2, or max_inner_product)
efc	int	150	HNSW construction time/accuracy trade-off
m	int	12	HNSW max number of connections per element
search_ef	int	40	HNSW search time/accuracy trade-off
embedder	Optional[Embedder]	OpenAIEmbedder()	Embedder instance for creating embeddings