Agno

User Guide

Examples

Reference

Changelog

FAQs

Discord

Community

Website

Website Knowledge Base

**Agno is a lightweight framework for building multi-modal Agents.**

Welcome to Agno

Your first Agent

**Agno provides a beautiful Agent UI for interacting with your agents.**

Agent Playground

Agent Observability

Install & Setup

Run Local Agent API

Contributing to Agno

Migrate from Phidata to Agno

Introduction

Product updates

Environment Variables Setup

Setting Environment Variables

TPM rate limiting

Tokens-per-minute rate limiting

Command line authentication

Connecting to Tableplus

Docker Connection Error

Could Not Connect To Docker

OpenAI Key Request While Using Other Models

Agent.run()

Structured Output

Multimodal Agents

Prompts

Tools

Knowledge

Memory

Storage

Reasoning

Teams

Compatibility

OpenAI

OpenAI Like

Anthropic Claude

AWS Bedrock Claude

Azure OpenAI

Cohere

DeepSeek

Fireworks

Gemini - AI Studio

Gemini - VertexAI

Groq

HuggingFace

Mistral

Nvidia

Ollama

OpenRouter

Sambanova

Together

Writing your own Toolkit

Python Functions

ArXiv

ArXiv Knowledge Base

Combined

Combined KnowledgeBase

CSV Knowledge Base

CSV URLs

CSV URL Knowledge Base

Docx

Docx Knowledge Base

Document

Document Knowledge Base

JSON

JSON Knowledge Base

LangChain

LangChain Knowledge Base

LlamaIndex

LlamaIndex Knowledge Base

PDF Knowledge Base

PDF URLs

PDF URL Knowledge Base

S3 PDF

S3 PDF Knowledge Base

S3 Text

S3 Text Knowledge Base

Text

Text Knowledge Base

Wikipedia

Wikipedia KnowledgeBase

Youtube

Youtube KnowledgeBase

Fixed Size Chunking

Agentic Chunking

Semantic Chunking

Recursive Chunking

Document Chunking

Cassandra

Cassandra Agent Knowledge

ChromaDB

ChromaDB Agent Knowledge

Clickhouse

Clickhouse Agent Knowledge

LanceDB

LanceDB Agent Knowledge

Milvus

Milvus Agent Knowledge

PgVector

PgVector Agent Knowledge

Pinecone

Pinecone Agent Knowledge

Qdrant

Qdrant Agent Knowledge

SingleStore

SingleStore Agent Knowledge

MongoDB

MongoDB Agent Knowledge

Postgres

Postgres Agent Storage

Sqlite

Sqlite Agent Storage

Singlestore

Singlestore Agent Storage

DynamoDB

DynamoDB Agent Storage

JSON Agent Storage

YAML

YAML Agent Storage

OpenAI Embedder

Gemini

Gemini Embedder

Ollama Embedder

Azure OpenAI Embedder

Cohere Embedder

Fireworks Embedder

HuggingFace Embedder

Mistral Embedder

FastEmbed

Qdrant FastEmbed Embedder

SentenceTransformers

SentenceTransformers Embedder

Together Embedder

Voyage AI

Voyage AI Embedder

**Use the `session_state` to cache intermediate results in a database.**

Caching

Advanced

Basic Agent

Agent with Tools

Agent with Knowledge

Agent with Storage

Agent Team

Custom Tools

Research Agent

Research Workflow

Image Agent

Image Generation

Video Generation

Audio Agent

Agent State

Agent Context

Agent Session

User Memories

Retry Functions

Human in the Loop

Finance Agent

Youtube Agent

Research Agent using Exa

Teaching Assistant

Recipe Creator

Movie Recommender

Books Recommender

Travel Agent

Blog Post Generator

Investment Report Generator

Personalized Email Generator

Startup Idea Validator

Agentic RAG

SQL Agent

Game Generator

GeoBuddy

Chess Battle

Audio Input Output

Multi-turn Audio Agent

Generate Images with Intermediate Steps

Generate Video using Models Lab

Generate Video using Replicate

Image to Audio Agent

Image to Image Agent

Video Caption Agent

Video to Shorts Agent

Traditional RAG with PgVector

Traditional RAG with LanceDB

Agentic RAG with PgVector

Agentic RAG with LanceDB

Agentic RAG with Reranking

Agentic RAG with Agent UI

RAG with LanceDB and SQLite

Combined Knowledge Base

DOCX Knowledge Base

Built-in Memory

Parameter	Type	Default	Description
`urls`	`List[str]`	`[]`	List of URLs to scrape
`reader`	`Optional[WebsiteReader]`	`None`	Reader to scrape website content
`max_depth`	`int`	`3`	Maximum depth of links to follow when scraping
`max_links`	`int`	`10`	Maximum number of links to follow when scraping