MervinPraison/PraisonAI

What is it?

Key features

• 01Single and multi-agent teams with handoffs between agents
• 02MCP tool integration over stdio, HTTP, WebSocket and SSE
• 03Persistent memory, sessions and knowledge retrieval (RAG)
• 04Workflow patterns including routing, parallel execution and loops
• 05Tools can run in Docker, E2B, Modal and other sandboxes
• 06Support for 100+ LLM providers, including local models

Use cases

• Automating research and report writing across multiple sources
• Building customer support bots on Telegram, Slack or Discord with memory
• Orchestrating coding agents and multi-step code generation tasks
• Creating data pipeline agents that query APIs and databases

Quick start

1. 1

   Install the lightweight core SDK

   pip install praisonaiagents
2. 2

   Set your LLM provider API key (the example uses OpenAI)

   export OPENAI_API_KEY="your-api-key"
3. 3

   Run the first agent from a Python script using the Agent class

Things to know

• !The quick start needs an API key from a paid LLM provider such as OpenAI.
• !Advanced sandbox features such as Docker, E2B or Modal require their own accounts or infrastructure.

Topics

FAQ

What is PraisonAI?

Python framework for building autonomous AI agents that research, plan, code and run tasks, with memory, RAG and 100+ LLM providers. PraisonAI is a framework for building AI agents that can research, plan, write code and execute multi-step tasks. It ships a core Python SDK (praisonaiagents) for defining agents with instructions, tools, memory and knowledge, plus a CLI, a YAML mode for running agent teams without code, and optional packages for a chat UI, a Langflow-based visual flow builder and a Telegram, Slack, Discord and WhatsApp dashboard.

Is PraisonAI free to use? What license does it have?

PraisonAI is open source under the MIT license. Check the license terms before using it in commercial products.

Is PraisonAI still maintained?

Actively maintained. The last commit to PraisonAI was on Oct 20, 2026.

What can I use PraisonAI for?

Automating research and report writing across multiple sources Building customer support bots on Telegram, Slack or Discord with memory Orchestrating coding agents and multi-step code generation tasks Creating data pipeline agents that query APIs and databases

How do I install PraisonAI?

Install the lightweight core SDK: pip install praisonaiagents Set your LLM provider API key (the example uses OpenAI): export OPENAI_API_KEY="your-api-key" Run the first agent from a Python script using the Agent class

Where can I get PraisonAI?

The source code and installation instructions are on GitHub: https://github.com/MervinPraison/PraisonAI