Framework · Agno Inc.
Agno at a glance
A Python framework and runtime: the SDK builds agents, teams and workflows, and AgentOS serves them as a FastAPI app on your own infrastructure. Sessions, memory and traces stay in your database, and a hosted Control Plane UI connects from the browser.
Best for: Teams that want to own a complete agent platform, with RBAC and Slack or WhatsApp interfaces, on their own servers.
Key facts
- Maintainer
- Agno Inc.
- Languages
- Python
- Licence
- Apache-2.0
- Latest version
3.1.0· October 1, 2026- Repository
- agno-agi/agno
- Hosted option
- AgentOS (self-hosted runtime) + Control Plane UI at os.agno.com
Quickstart
Install the package, set the credentials, then run the smallest agent from the official docs. Model IDs are the ones the docs use and may not be enabled on your account.
uv pip install -U agno openaifrom agno.agent import Agent
from agno.models.openai import OpenAIResponses
agent = Agent(
name="Quickstart Agent",
model=OpenAIResponses(id="gpt-5.2"),
)
if __name__ == "__main__":
agent.print_response(
"Say hello and introduce yourself in one sentence.", stream=True
)Features
| Any model | Yes | |
|---|---|---|
| MCP client | Yes | |
| MCP server | Yes | AgentOS serves your platform as an MCP server |
| Multi-agent | Yes | Agents, teams and workflows |
| Durable runs | Yes | AgentOS stores sessions and memory in your own database |
| Human approval | Yes | Pause runs for confirmation; gate tools behind admin approval |
| Tracing | Yes | OpenTelemetry tracing, run history and audit logs |
| Evals | Not confirmed |
Recent changes
- The 3.x line repositions Agno as a framework and runtime for agent platforms: SDK, AgentOS and Control Plane.
- Onboarding now starts from coding-agent prompts and agentos-* starter repos.