Agno
High-performance, model-agnostic Python framework for building, running, and managing production AI agents at scale.
Agno
Agno (formerly Phidata) is a production-grade Python framework for building multi-modal, multi-provider AI agents. It benchmarks at ~2 µs agent creation time and ~3.75 KiB memory per agent — making it one of the fastest agent runtimes available.
Key Features
- Model-agnostic — works with OpenAI, Anthropic, Google Gemini, Groq, Cohere, local Ollama, and 40+ providers
- Multi-modal — agents natively handle text, image, audio, and video inputs
- Built-in memory — short and long-term memory via SQLite, PostgreSQL, or Redis
- Knowledge bases — attach structured knowledge from PDFs, URLs, databases, and vector stores
- Teams of agents — coordinate multiple agents in parallel or sequential workflows
- Monitoring — first-class integration with Agno's own monitoring dashboard and OpenTelemetry
- Minimal boilerplate — define a complete agent with tools in ~10 lines
Quick Start
pip install agno
from agno.agent import Agent
from agno.models.openai import OpenAIChat
from agno.tools.duckduckgo import DuckDuckGoTools
agent = Agent(
model=OpenAIChat(id="gpt-4o"),
tools=[DuckDuckGoTools()],
markdown=True,
)
agent.print_response("What are the top AI papers of 2026?", stream=True)
Install via ai-supply
npx ai-supply add agno-agent-framework
Curated mirror of the open-source Agno (MPL-2.0). Get it from the source.
Compromise signals — malicious or tampered code (leaked secrets, backdoors, a dropped executable) — reduce the score, and known dependency CVEs carry a bounded penalty (they warrant review but never QUARANTINE — update the dependency to clear). Other dangerous-by-capability traits are risk surface, expected for some capabilities. Every finding is mapped to its OWASP control below.
Findings mapped to the OWASP Top 10 for LLM Applications (2025) and the OWASP Machine Learning Security Top 10. Expand any flagged control for the exact findings — compromise reduces the score; expected/risk-surface do not, except a known CVE, which carries a small bounded penalty (high/critical → Review).
The same gate an agent runs before installing (POST /api/v1/trust/agno-agent-framework/check). Click a policy:
Consume Agno programmatically. Authenticate with an API key or session — see Authorize an agent.
# Agents: CHECK BEFORE YOU INSTALL (no auth) — score, grade, level, capability manifest
curl https://ai-supply.store/api/v1/trust/agno-agent-framework
# Gate against your org policy (returns { pass, violations })
curl -X POST https://ai-supply.store/api/v1/trust/agno-agent-framework/check \
-H "Content-Type: application/json" \
-d '{"minGrade":"B","denyPermissions":["shell"],"denyUnknownEgress":true}'
# CLI
npx ai-supply add agno-agent-framework
# REST (install → download)
curl -X POST https://ai-supply.store/api/v1/listings/agno-agent-framework/install \
-H "Authorization: Bearer $AIM_KEY"
# MCP tool
install_listing({ "slug": "agno-agent-framework" })OpenAPI spec →Curated mirror — latest upstream source. See the repository for tagged releases.