Blog ·
Running several VMs together on one Mac
How to create Windows, macOS, and Linux VMs from one manager screen, run them at the same time, and set them aside to pick up where you left off.
Manage them all from one screen
The Any OS manager screen gathers every VM on this Mac. Each VM shows its operating system, name, and state, and you can start it, open its screen, suspend it, change its settings, or save a checkpoint right there.
To create a new VM, click + and choose an operating system. The app downloads and installs it for you: Windows 11 from Microsoft, macOS from Apple's restore image, and Ubuntu from the official ARM64 image. The screen layout is covered in Managing your VM list.
Run them at once, each in its own window
Each VM runs in its own window. You can work on documents in Windows while a server runs in Ubuntu in the next window, and check another version of an app in a macOS VM. The screenshots in this post were taken on a MacBook Pro with an Apple M1 Max and 32 GB of memory, with Windows 11, macOS, and Ubuntu VMs running together.
Closing a window keeps the VM running; reopen it with Open Screen on the manager screen. To show only guest app windows on your Mac, turn on Seamless mode.
Split CPU and memory between VMs
Set CPU cores, memory, and disk size separately for each VM. A single VM can get at most your Mac's core count minus 2, and 75% of your Mac's memory. If the memory of the running VMs would add up to more than your Mac has, Any OS doesn't start the new VM and asks you to shut down or suspend another one.
Recommended values and ranges are in CPU, memory, and disk.
Set aside the VMs you aren't using
| Suspend | |
|---|---|
| Saved to | Mac disk |
| Mac memory | Given back |
Suspend saves your open apps and windows to disk and gives the memory back to your Mac. It's a good fit when another VM needs more memory. Details are in Suspend and resume.
Run several experiments from the same setup
A checkpoint stores a shut-down VM's disk and boot state as a whole. Restoring it leaves the original VM as it is and creates a new VM from that moment, so you can branch several experiments from the same starting point.
To split VMs between AI agents and your own use, see Handing a VM to AI agents.