Run agents on Valkey for fast in-memory storage

ValkeyDb brings Valkey to Agno as an in-memory database for agents, teams, and workflows. Your sessions and state live in memory, so reads and writes stay quick under load.

from agno.agent import Agent
from agno.db.valkey import ValkeyDb
 
db = ValkeyDb(host="localhost", port=6379)
agent = Agent(db=db)

Not every agent needs a full relational database sitting behind it just to hold session state. When latency matters more than durable long-term storage, backing your agents with Valkey keeps that path light and fast. Point an agent, a team, or a workflow at it and the storage layer stops being the slow part.

See the Valkey database docs for more.