Praison AI

PraisonAI is a production-ready Multi-AI Agents framework with self-reflection, designed to create AI Agents to automate and solve problems ranging from simple tasks to complex challenges. It provides a low-code solution to streamline the building and management of multi-agent LLM systems, emphasising simplicity, customisation, and effective human-agent collaboration.

Build production-ready AI agents that reason, remember, and act autonomously — with just a few lines of code.

---

Why PraisonAI?

Runs Anywhere

Your machine, cloud, or edge. Self-hosted with full control over your data.

100+ LLM Models

OpenAI, Anthropic, Google, Ollama, Groq — seamlessly switch between any provider.

Memory & Knowledge

Persistent memory, RAG knowledge bases, and context-aware conversations.

Self-Reflection

Agents that evaluate and improve their own responses for higher accuracy.

140+ Built-in Tools

Web search, file operations, databases, APIs — all ready to use out of the box.

Multi-Agent Workflows

Sequential, parallel, hierarchical orchestration with autonomous coordination.

---

Differentiators

🧠 Self-Reflection & Reasoning

Unlike simple chatbots, PraisonAI agents can reflect on their outputs and iteratively improve them. Built-in reasoning capabilities enable multi-step problem solving without manual prompting.

agent = Agent(
    name="analyst",
    self_reflect=True,      # Enable self-correction
    reasoning=True,         # Enable chain-of-thought
    max_reflect=3           # Up to 3 reflection cycles
)

💾 Persistent Memory

Agents remember past interactions across sessions. Short-term, long-term, and episodic memory systems keep context without re-prompting.

agent = Agent(
    name="assistant",
    memory=True,             # Enable persistent memory
    user_id="user_123"       # Scoped to specific user
)

📚 Knowledge Bases (RAG)

Connect agents to your documents, databases, and APIs. Built-in RAG with auto-chunking, embeddings, and semantic search.

agent = Agent(
    name="researcher",
    knowledge="./docs/",       # Local documents
    knowledge_sources=[        # Or multiple sources
        "https://example.com/api",
        "./pdfs/"
    ]
)

🌐 Browser Control

Agents can navigate websites, fill forms, click buttons, and extract data — fully automated browser automation.

agent = Agent(
    name="web_agent",
    tools=["browser_tool"],   # Built-in browser
    llm="gpt-4o"
)
agent.run("Book a flight from NYC to LA for next Tuesday")

💬 11+ Messaging Platforms

Deploy agents to Slack, Discord, Telegram, WhatsApp, Signal, LINE, iMessage, and more — with a single command.

praisonai bot slack --token $SLACK_BOT_TOKEN --app-token $SLACK_APP_TOKEN

🔌 Extensible Plugins

Write custom tools in Python or use any of the 140+ pre-built integrations. MCP protocol support for external tool servers.

from praisonaiagents import Agent

def custom_tool(query: str) -> str:
    """My custom tool description."""
    return f"Result for {query}"

agent = Agent(tools=[custom_tool])

---

Quick Start

No Code

1. Install

   pip install praisonai
   
2. Set API Key

   export OPENAI_API_KEY=sk-xxx
   
3. Run

   praisonai --auto "Research the top 5 AI trends in 2025 and create a report"
   

   That’s it! Agents are automatically created and orchestrated.

Python SDK

1. Install

   pip install praisonaiagents
   
2. Create Agent

   from praisonaiagents import Agent
   
   agent = Agent(
       name="researcher",
       instructions="You are a research analyst specializing in AI trends.",
       llm="gpt-4o"
   )
   
   result = agent.start("What are the top AI trends for 2025?")
   print(result)
   

Multi-Agent

from praisonaiagents import Agent, Task, AgentTeam

# Create specialized agents
researcher = Agent(
    name="Researcher",
    role="Research Analyst",
    goal="Find accurate, up-to-date information",
    llm="gpt-4o"
)

writer = Agent(
    name="Writer",
    role="Content Writer",
    goal="Create engaging, clear content",
    llm="gpt-4o"
)

# Define tasks
research_task = Task(
    name="research",
    description="Research AI trends for 2025",
    agent=researcher
)

write_task = Task(
    name="write",
    description="Write a blog post based on the research",
    agent=writer
)

# Orchestrate
agents = AgentTeam(
    agents=[researcher, writer],
    tasks=[research_task, write_task],
    process="sequential"
)

result = agents.start()

---

Platform Integrations

Slack

Deploy AI bots to Slack workspaces

Discord

Create Discord bots with agent intelligence

Telegram

Build Telegram bots in minutes

WhatsApp

Send messages via WhatsApp Business

GitHub

Automate repos, issues, and PRs

Google

Calendar, Sheets, Drive, Gmail

Notion

Read and write Notion pages

Jira

Manage Jira issues and projects

---

CLI Commands

Full-featured CLI for deploying and managing AI agents:

Messaging Bots

# Start a Telegram bot
praisonai bot telegram --token $TELEGRAM_BOT_TOKEN

# Slack with full capabilities
praisonai bot slack --token $SLACK_BOT_TOKEN --app-token $SLACK_APP_TOKEN \
  --agent agents.yaml --browser --web --memory

# Discord with custom tools
praisonai bot discord --token $DISCORD_BOT_TOKEN \
  --tools DuckDuckGoTool WikipediaTool

Browser Control

# Navigate and interact with websites
praisonai browser navigate https://example.com
praisonai browser run "Click the submit button"
praisonai browser screenshot --output page.png
praisonai browser doctor

Skills & Plugins

# Manage skills
praisonai skills list
praisonai skills check --verbose
praisonai skills create --name csv-analyzer --description "Analyze CSV" --script

# Manage plugins
praisonai plugins list --enabled
praisonai plugins doctor

Sandbox

# Secure execution
praisonai "Run code" --sandbox strict
praisonai sandbox status
praisonai sandbox explain --agent work

Bot CLI

Deploy messaging bots

Browser CLI

Browser automation

Skills CLI

Manage agent skills

Plugins CLI

Plugin management

---

Architecture

---

Use Cases

Customer Support

Build intelligent support agents that resolve issues autonomously, escalate when needed, and learn from interactions.

Research & Analysis

Create agents that gather data, analyze trends, and produce detailed reports with citations.

Content Creation

Deploy agents that research, write, edit, and optimize content across formats — blogs, docs, social media.

Process Automation

Automate complex workflows with multi-agent coordination — from data pipelines to deployment.

Code Generation

Agents that understand your codebase, write tests, fix bugs, and implement features.

Personal Assistant

Build assistants that manage calendars, emails, reminders, and integrate with your tools.

---

Explore

Agents

Create and configure intelligent agents

Learn more

Tools

140+ built-in tools and integrations

Browse tools

Memory

Persistent memory and knowledge bases

Set up memory

MCP

Model Context Protocol integrations

Connect services

Messaging Bots

Slack, Discord, Telegram, and more

Deploy bots

API Reference

Full SDK documentation

View API

---