Skip to main content
from praisonaiagents import Agent, tool

@tool
def hello(name: str) -> str:
    """Greet someone by name."""
    return f"Hello, {name}!"

agent = Agent(name="Demo", tools=[hello])
agent.start("Say hello to Alex using the hello tool")
The user prototypes a custom tool with the Tools Creator GPT, then runs it through a PraisonAI agent.

​
PraisonAI Tools Creator GPT

Use PraisonAI Tools Creator GPT to get started quickly.

​
Quick Start

1

Set API key

export OPENAI_API_KEY=your_openai_api_key
2

Create an agent

from praisonaiagents import Agent

agent = Agent(
    name="GPTAgent",
    instructions="Answer questions accurately and concisely.",
    llm="gpt-4o",
)

agent.start("Explain quantum entanglement in simple terms")

​
Best Practices

GPT-4o handles multi-step reasoning and tool use better than smaller models.
Use lower temperature (0.1-0.3) for factual tasks and higher (0.7-0.9) for creative tasks.
Enable JSON mode when you need parseable, structured output from the agent.
Route simpler subtasks to GPT-4o-mini to reduce costs while keeping quality high.

Custom Tools

Build your own agent tools

Tools Overview

Browse PraisonAI tool documentation