Database

Persist sessions and connect Agno features to database-backed storage.

Set db on an agent, team, or workflow to persist its sessions and runs. Database implementations can also back memories, learnings, evaluations, knowledge content, traces, and schedules. Supported tables vary by provider.

  • Chat history. Include previous messages in context for multi-turn conversations.
  • Session persistence. Store session information and conversation history across requests.
  • State management. Store session state across runs.
  • Context control. Store session summaries and the runs used to build model context.
  • Memory and knowledge. Store user memories, learned knowledge, and knowledge content metadata.
  • Tracing and evaluation. Store detailed traces for debugging, monitoring, and building evaluation datasets.
  • Data access. Query records in the database you configure and use them to build evaluation datasets or review run quality.

Prerequisites

For the SQLite/OpenAI examples:

uv pip install -U "agno[openai,sqlite]"
export OPENAI_API_KEY="your_openai_api_key"

Quick Start

from agno.agent import Agent
from agno.db.sqlite import SqliteDb

agent = Agent(
    db=SqliteDb(db_file="agent.db"),
    add_history_to_context=True,
    num_history_runs=3,
)

# First message
agent.print_response("I'm working on a Python API project", session_id="dev_session")

# Later, the agent remembers the context
agent.print_response("What testing framework should I use?", session_id="dev_session")

The agent now persists sessions and includes the last 3 runs in every request.

Database storage overview

Guides

Chat History

Include previous messages in context for multi-turn conversations.

Session Storage

Store and retrieve session data from your database.

Session Summaries

Condense long conversations to manage token costs.

Storage Control

Choose what gets persisted to your database.

Works With Teams and Workflows

Agents, teams, and workflows all accept the db parameter:

The PostgreSQL fragment requires a running PostgreSQL database and its connection URL. Install the driver with uv pip install -U "psycopg[binary]", then replace the sample credentials below.

from agno.agent import Agent
from agno.team import Team
from agno.workflow import Workflow
from agno.db.postgres import PostgresDb

db = PostgresDb(db_url="postgresql+psycopg://user:pass@localhost:5432/mydb")

team = Team(members=[Agent(name="Assistant")], db=db)
workflow = Workflow(db=db)

Supported Databases

Use SQLite for local development or choose a networked provider for a deployed application. See the database index.

Async Support

For async applications, use the async database classes:

from agno.agent import Agent
from agno.db.postgres import AsyncPostgresDb

agent = Agent(
    db=AsyncPostgresDb(db_url="postgresql+psycopg_async://..."),
)

Troubleshooting