Agent-as-a-Service Platform
+12 Monkeys

+12
MONKEYS

Build and deploy production-ready AI agents
in under 10 minutes. No code required.

Launch Agent Wizard

Scroll to learn more

01 / PLATFORM

From conversation to deployment.
In minutes, not months.

+12 Monkeys is the open-source platform that turns natural language into production-ready AI agent systems. Describe what you need, and receive a complete deployable package — Python, TypeScript, Rust, or Go.

Built on the Model Context Protocol (MCP) standard. 20 agent templates. 8 frameworks. 4 languages. Zero boilerplate.

Pillar 01

CONFIGURE

[ Chat . Define . Architect ]

Describe your agent in natural language. The wizard extracts requirements, recommends frameworks, selects MCP servers, and architects your system — all through conversation.

Start Building

mcp-config.json

{"mcpServers": {"research-agent": {"command": "python","args": ["agent.py"],"env": {"OPENAI_API_KEY": "sk-..."}}}}

Tell the wizard what you need in plain English. It generates your MCP configuration, picks the right servers, and wires everything together automatically.

agent.py — generated

from langgraph.graph import StateGraphfrom mcp import ClientSessionclass ResearchAgent:def __init__(self): self.graph = StateGraph(AgentState) self.graph.add_node("search", self.search) self.graph.add_node("analyze", self.analyze) self.graph.add_node("summarize", self.summarize) self.graph.add_edge("search", "analyze") self.graph.add_edge("analyze", "summarize")async def run(self, query: str):async with ClientSession() as mcp:return await self.graph.ainvoke({"query": query})

Production-ready agent code with MCP integration, Dockerfile, docker-compose, and environment configs — all generated from your conversation.

Pillar 02

GENERATE

[ Python . TypeScript . Rust . Go ]

Complete deployable packages in seconds. Agent code, Dockerfile, docker-compose, environment configs, MCP server setup, Kubernetes manifests, and AWS SAM templates — all generated.

Generate Package
Pillar 03

DEPLOY

[ Local . Cloud . Export ]

One command to production. Docker Compose locally, Railway / Render / Vercel in the cloud, or export and self-host on your own infrastructure. 30 seconds to running.

Deploy Now

terminal

$ docker compose up -d Container agent-mcp    Started Container agent-api    Started Container agent-worker Started$ railway up Deploying to production... Build completed in 28s Live at agent.railway.app# Or export & self-host anywhere$ +12m export --format k8s Kubernetes manifests written

Run locally with Docker, deploy to Railway/Vercel with one command, or export Kubernetes manifests for your own infra. Your choice.

20 TEMPLATES

Production-ready agent architectures for every domain. Customer service, research, healthcare, finance, defense, and more.

Research

Research Agent

Web search, data extraction,
academic papers, summarization.
LangGraph workflow engine.

Enterprise

Multi-Agent Team

Manager, Executor, Critic, Planner.
Collaborative problem-solving.
AutoGen coordination.

Operations

Customer Service

FAQ, ticket routing, sentiment.
Slack + Email + CRM integration.
CrewAI multi-agent system.

Open Source · MCP Native

Deploy your first agent.
Production-ready in under 10 minutes.

+12 Monkeys © 2025. MIT License. All Rights Reserved.