Run Moltbot in a macOS VM with Lume

January 27, 20265 min read#AI, #automation, #macos, #vm, #lume

molt

What is Moltbot?

Moltbot (formerly Clawdbot) is an open-source personal AI assistant that runs entirely on your own machine. Unlike cloud-based AI services, Moltbot keeps your data private by default and gives you full control over your AI assistant.

Key Features

  • Multi-platform messaging — Works with WhatsApp, Telegram, Discord, Slack, Signal, and iMessage
  • Persistent memory — Learns your preferences and context over time
  • Web browsing — Can browse, fill forms, and interact with websites
  • File management — Handles files and executes shell commands
  • Extensible — 50+ integrations including Claude, GPT, Spotify, Philips Hue, Obsidian, Twitter, Gmail, and GitHub
  • Self-modifying — Can write its own extensions and skills

Why Run Moltbot in a VM?

Running Moltbot in a virtual machine provides several advantages:

  • Complete isolation — Your host system stays protected
  • Easy reset — Something breaks? Clone from backup and you’re back
  • Headless operation — Use your Mac normally while Moltbot runs in the background
  • Portable — Clone and move VMs between machines
  • iMessage support — Unlike Linux, macOS VMs give you access to iMessage integration

Prerequisites

  • Apple Silicon Mac (M1, M2, M3, M4 — any will work)
  • macOS on your host machine
  • At least 16GB RAM recommended (8GB minimum)
  • 50GB+ free disk space for the VM

Quick Setup Guide

1. Install Lume

curl -fsSL https://raw.githubusercontent.com/trycua/cua/main/libs/lume/scripts/install.sh | bash

Add ~/.local/bin to your PATH if not already present:

echo 'export PATH="$HOME/.local/bin:$PATH"' >> ~/.zshrc
source ~/.zshrc

2. Create a macOS VM

lume create macos --name moltbot-vm --unattended tahoe

This downloads macOS and creates the VM, automatically complete the Setup Assistant using VNC connection and OCR. After this process, your VM is ready to use with the user lume, password lume

3. Run the VM Headlessly

Once setup is complete, run without display:

lume run moltbot-vm --no-display

4. SSH into the VM

Run this command to get IP address of the VM

lume get moltbot-vm
ssh lume@198.164.x.x

Replace lume with the username you created during setup if you set it up manually.

5. Install Moltbot Inside the VM

curl -fsSL https://molt.bot/install.sh | bash

Or via npm:

npm i -g clawdbot
clawdbot onboard

6. Configure Your Channels

Edit ~/.clawdbot/clawdbot.json to add your messaging channels:

{
  "channels": {
    "whatsapp": { "enabled": true },
    "telegram": { "enabled": true, "token": "YOUR_BOT_TOKEN" },
    "discord": { "enabled": true, "token": "YOUR_BOT_TOKEN" }
  }
}

7. Login and Start

clawdbot login
clawdbot start

The gateway runs automatically via the daemon you installed.

6. Setup Xcode

if you are an iOS developer like me, you can now install Xcode and other relevant tools for iOS development.

ios

After that, I can ask my bot to generate code, run the app, and send me the screenshot of the running app on Telegram, very cool 😎

telegram

VM Management Tips

Save a Golden Image

Before heavy customization, snapshot it:

lume stop moltbot-vm
lume clone moltbot-vm moltbot-vm-backup

Reset Anytime

lume delete moltbot-vm
lume clone moltbot-vm-backup moltbot-vm

Keep It Running 24/7

If using a laptop, keep it plugged in. Configure Energy Saver to prevent sleep, or the VM will pause when your Mac sleeps.

FAQ

Can I use any Mac, not just Mac Mini? Yes. MacBook Air, MacBook Pro, iMac, Mac Studio — any Apple Silicon Mac works.

Does it work on M1? Yes. All Apple Silicon works — M1, M2, M3, M4. That dusty MacBook Air from 2020? Yep.

What about cloud? Lume works on AWS EC2 dedicated Mac hosts too. But at ~$720/month, your home Mac is already paid for.

Resources

Quick Drop logo

Profile picture

Personal blog by An Tran. I'm focusing on creating useful apps.
#Swift #Kotlin #Mobile #MachineLearning #Minimalist


© An Tran - 2026