Deploy

Deploy PraisonAI agents to cloud, Docker, or as local servers

PraisonAI offers two deployment approaches: Deploy to Cloud/Docker using praisonai deploy CLI, or Run as a Server locally using Python SDK.

Choose Your Path

Deploy to Cloud

Deploy agents to AWS, Azure, GCP, or Docker

Run as Server

Start HTTP, MCP, A2A, or AGUI servers locally

---

Deploy to Cloud

Use praisonai deploy to deploy agents to cloud providers or Docker containers.

The auto-generated api.py (produced by praisonai deploy run --type api) imports bleach to sanitise markdown output before returning it as HTML. Install it alongside praisonai:

pip install praisonai bleach

Without bleach, the generated server fails at startup with ModuleNotFoundError: No module named 'bleach'.

Local API

pip install praisonai bleach
praisonai deploy run --type api
# As of 4.6.34, requires bearer token by default
# See API Server Authentication guide

Docker

praisonai deploy run --type docker

AWS

praisonai deploy run --type cloud --provider aws

GCP

praisonai deploy run --type cloud --provider gcp

Azure

praisonai deploy run --type cloud --provider azure

Cloud Deployment Pages

Target	Description	Guide
API Server	Local FastAPI server for development	API Guide · Auth Guide
Docker	Containerized deployment	Docker Guide
AWS	Deploy to AWS ECS/Fargate	AWS Guide
Azure	Deploy to Azure Container Apps	Azure Guide
GCP	Deploy to Google Cloud Run	GCP Guide

Deploy CLI Commands

Command	Description
praisonai deploy run	Execute deployment
praisonai deploy init	Generate agents.yaml with deploy config
praisonai deploy doctor	Check deployment readiness
praisonai deploy status	View deployment status
praisonai deploy destroy	Remove deployment

Deploy CLI Reference

Complete guide to all praisonai deploy commands and options

---

Run as Server

Start agents as servers locally using Python SDK or CLI.

HTTP Server

from praisonaiagents import Agent

agent = Agent(instructions="You are helpful")
agent.launch(port=8000)

MCP Server

from praisonaiagents import ToolsMCPServer

server = ToolsMCPServer(name="my-tools")
server.run(transport="sse", port=8080)

CLI

praisonai serve agents --port 8000

All Serve Commands

All server commands are now unified under praisonai serve <type>. Run praisonai serve to see all available options.

Use praisonai serve <type> to start any server. All options in one place:

Command	Protocol	Default Port	Description
praisonai serve agents	HTTP	8000	Agents as HTTP REST API
praisonai serve gateway	WebSocket	8765	Multi-agent real-time coordination
praisonai serve mcp	STDIO/SSE	8080	MCP server for Claude/Cursor
praisonai serve acp	STDIO	-	Agent Client Protocol for IDEs
praisonai serve lsp	STDIO	-	Language Server Protocol
praisonai serve ui	HTTP	8082	Chainlit web interface
praisonai serve rag	HTTP	9000	RAG query server
praisonai serve registry	HTTP	7777	Package registry server
praisonai serve docs	HTTP	3000	Documentation preview
praisonai serve scheduler	Background	-	Job scheduler daemon
praisonai serve recipe	HTTP	8765	Recipe runner server
praisonai serve a2a	JSON-RPC	8001	Agent-to-Agent protocol
praisonai serve a2u	SSE	8002	Agent-to-User event stream
praisonai serve unified	HTTP/SSE	8765	All providers combined
praisonai serve openai	HTTP/SSE	8765	OpenAI API compatibility layer

Unified Serve Reference

Complete guide to all praisonai serve commands and options

Server Type Guides

Server	Use Case	Guide
Unified Serve	All serve commands in one place	Serve Commands
Agents	HTTP REST API for single/multi-agent systems	Agents Server
Gateway	WebSocket multi-agent real-time coordination	Gateway Server
MCP	Expose tools to Claude Desktop, Cursor	MCP Server
A2A	Agent-to-Agent protocol communication	A2A Server
AGUI	CopilotKit frontend integration	AGUI Server

---

API Reference

Documentation for all server endpoints.

API	Protocol	Reference
Agents API	HTTP REST	Endpoints
MCP API	MCP Protocol	Endpoints
A2A API	JSON-RPC	Endpoints
AGUI API	SSE Streaming	Endpoints

---

Quick Decision Guide

I want to…	Use
Test locally	praisonai deploy run --type api or agent.launch() (auto-generated API server prints bearer token on each startup)
Deploy to production	praisonai deploy run --type cloud --provider <aws/gcp/azure>
Containerize my agent	praisonai deploy run --type docker
Integrate with Claude Desktop	Tools MCP Server
Build a chat UI	AGUI Server with CopilotKit
Connect agents together	A2A Server