PraisonAI Quick Start: Building Your First Multi-Agent System

Getting Started in Minutes

PraisonAI lets you go from zero to a working multi-agent system in just a few minutes. This quick start guide covers installation, configuration, and your first agent team in action.

Step 1: Install

pip install praisonai

That is it. One command and you have the full PraisonAI framework ready to use.

Step 2: Configure Your Agent Team

Define agents in a simple YAML config. Each agent has a name, role, goal, backstory, and LLM assignment.

from praisonai import PraisonAI

agents = PraisonAI(
    agents_config="""
    - name: Researcher
      role: Research Specialist
      goal: Find accurate and up-to-date information
      backstory: Expert researcher with web access
      llm: gpt-4
    - name: Writer
      role: Content Writer
      goal: Write compelling, well-structured content
      backstory: Skilled writer who creates engaging content
      llm: gpt-4
    """,
    tasks=["Research and write an article about AI agents"]
)

agents.start()

Step 3: Run with Docker

For production deployments, use the Docker setup from the repo:

git clone https://github.com/MervinPraison/PraisonAI.git
cd PraisonAI/docker
docker compose up -d

This starts the web UI, API server, and agent workers. Access the dashboard at http://localhost:8080.

Supported LLMs

Set your API keys as environment variables and agents auto-detect them. Supported providers include OpenAI, Anthropic, Google Gemini, DeepSeek, Ollama (local), Groq, Azure, and many more. For complete privacy, run everything locally with Ollama.

Advanced: Adding Memory

Enable Mem0 for persistent agent memory. Agents will remember past conversations and improve over time:

agents = PraisonAI(
    agents_config=config,
    tasks=["Continue our research from yesterday"],
    memory=True
)

Monitoring

The AgentFlow view in the dashboard shows how tasks propagate through your agent network in real time, making it easy to debug and optimize your workflows.

Conclusion

With just a few lines of Python, you can deploy a sophisticated multi-agent system. PraisonAI abstracts away the complexity of agent orchestration, letting you focus on what your agents should accomplish rather than how to wire them together.