Agno

By Agno

High-performance framework and runtime (AgentOS) for multi-agent systems

Website ↗ Repository ↗ Docs ↗

v2.8.02026-07-20PythonApache-2.0

#multi-agent #runtime #memory #python

41.4k Open source Free tier Paid

Doc version 1.0 · Updated June 28, 2026

Related concepts

Harness Engineering Doc version 1.0 · Updated June 28, 2026 Wrapping an LLM in scaffolding — a control loop, tools, a sandbox, guardrails, and evaluation — so it behaves reliably in production.

Related tools

CrewAI

Multi-Agent Orchestration

LangGraph

Multi-Agent Orchestration

Overview

Agno (formerly Phidata) is a high-performance Python framework for building agents with memory, knowledge, and tools, plus AgentOS — a runtime for serving and scaling them in production.
It is model-agnostic and emphasizes speed and keeping your data inside your own infrastructure.

flowchart LR
  prompt[prompt] --> agent[Agent]
  agent --> model[(model)]
  agent --> mem[memory / tools]
  agent --> out[response]

The Code samples tab shows a single agent backed by Claude.

When to use it

Choose Agno when you want a fast, batteries-included agent abstraction and an in-house runtime (AgentOS) to deploy multi-agent systems without sending data to a third-party platform.

Pricing

Open source (Apache-2.0) and free to self-host. The hosted AgentOS Control Plane is paid per seat and live connection.

Plan	Price	Notes
Free	$0	Build agent systems; Control Plane for a local AgentOS.
Pro	$150/mo	Live AgentOS Control Plane, 4 seats included (+$30/seat, +$95/live connection)
Enterprise	Custom	Dedicated support & SLA, SSO/RBAC, self-hosted Control Plane

Source ↗ · as of 2026-06-27

Example

Python · agent

When to use

Use Agno when you want one fast, model-agnostic agent abstraction with memory, knowledge, and tools — and a path to run it in production.

How it flows

An Agent pairs a model with optional memory, knowledge, and tools; print_response runs it and streams the reply.

Structure

flowchart LR
  prompt[prompt] --> agent[Agent]
  agent --> model[(model)]
  agent --> mem[memory / tools]
  agent --> out[response]

from agno.agent import Agent
from agno.models.anthropic import Claude

agent = Agent(model=Claude(id="claude-opus-4-8"), markdown=False)
agent.print_response("Say hello in one short sentence.")

Example

Agno — a single agent backed by Claude

A real, runnable mini-project. Download it and run with Docker.

View samples/agno_1 on GitHub ↗

June 28, 2026

Agno — a single agent backed by Claude

A tiny Agno sample: it builds one agent with a Claude model and runs your prompt.

Configure

cd samples/agno_1
cp .env.sample .env
# edit .env: set ANTHROPIC_API_KEY (and optionally MODEL)

Run with Docker

cd samples/agno_1
docker build -t aas-agno .
docker run --rm --env-file .env aas-agno "Say hello in one short sentence."

Run with Docker (in a devcontainer with DooD)

The foreground docker run may print nothing under Docker-outside-of-Docker — run detached and follow the logs:

docker logs -f "$(docker run -d --env-file .env aas-agno "Say hello in one short sentence.")"

Run locally

cd samples/agno_1
pip install -r requirements.txt
python app.py "Say hello in one short sentence."

Example run

Output varies by model and run — LLMs are non-deterministic. One run with claude-opus-4-8:

Hello, it's nice to meet you!

Files

.env.sample

# Copy to `.env` and add your Anthropic API key.
#   cp .env.sample .env
ANTHROPIC_API_KEY=

# Any Claude model id.
MODEL=claude-opus-4-8

app.py

"""Agno: a single agent backed by Claude (Anthropic).

Builds one Agno agent and runs a prompt. Model is configurable via MODEL.

Run:
    docker build -t aas-agno .
    docker run --rm --env-file .env aas-agno "Say hello in one short sentence."
"""

import os
import sys

from dotenv import load_dotenv
from agno.agent import Agent
from agno.models.anthropic import Claude

load_dotenv()

MODEL = os.environ.get("MODEL", "claude-opus-4-8")


def main() -> None:
    prompt = " ".join(sys.argv[1:]) or "Say hello in one short sentence."
    agent = Agent(model=Claude(id=MODEL), markdown=False)
    print(agent.run(prompt).content)


if __name__ == "__main__":
    main()

Dockerfile

FROM python:3.13-slim

ENV PYTHONUNBUFFERED=1

WORKDIR /app

COPY requirements.txt .
RUN pip install --no-cache-dir -r requirements.txt

COPY app.py .

ENTRYPOINT ["python", "app.py"]