Skip to main content
Import the SDK to drive agents from your own Python code, then reach for the praisonai wrapper for YAML-driven orchestration.
Pick the install extra that matches how you want to run the API.

Core Modules

praisonai

The main package containing core functionality.

praisonai.auto

Automated agent generation functionality.

praisonai.agents_generator

Framework-specific agent generation and orchestration:
  • CrewAI support (requires praisonai-frameworks[crewai])
  • AG2 (Formerly AutoGen) support (requires praisonai-frameworks[autogen])

praisonai.cli

Command-line interface with framework-specific handling.

praisonai.deploy

Deployment utilities.
As of PraisonAI 4.6.x, CloudDeployer (from the deploy.py module) was removed. Direct users of from praisonai.deploy import CloudDeployer should migrate to praisonai deploy CLI commands.

Version

praisonaiagents exposes the installed package version as __version__:

Installation Options

Framework-specific Features

CrewAI

When installing with pip install "praisonai-frameworks[crewai]", you get:
  • CrewAI framework support
  • PraisonAI tools integration
  • Task delegation capabilities
  • Sequential and parallel task execution

AG2 (Formerly AutoGen)

When installing with pip install "praisonai-frameworks[autogen]", you get:
  • AG2 framework support
  • PraisonAI tools integration
  • Multi-agent conversation capabilities
  • Code execution environment

How It Works

Your code calls the SDK or the praisonai wrapper, which generates agents and runs them against the chosen framework.

Best Practices

Use from praisonaiagents import Agent for embedding agents in Python. Reach for the praisonai wrapper only when you need YAML orchestration.
Add praisonai-frameworks[crewai] or [autogen] only when you target those frameworks — the core package stays lighter without them.
Use from praisonaiagents import __version__ to log the installed version and pin all packages to the same release cycle.
Use praisonai deploy run --type docker|cloud instead of the removed CloudDeployer class.

Installation

Choose the right package for your setup.

Quick Start

Build and run your first agent.