Five AI desktops: compare their approach before choosing one

Omarchy: shape the desktop around agents

Omarchy combines Arch Linux, Hyprland, coordinated themes and coding-agent workflows. The work-session timer mentioned in the video is NetworkChuck’s demonstration, not a performance test by Luke.

1. Open Omarchy’s official site and manual to inspect installation and hardware guidance.
2. Look at its window shortcuts and theme behaviour before deciding whether the keyboard-led approach suits you.
3. Keep paid Claude, Codex or other AI-service costs separate from the free operating system.

deepin: AI beside your reading and writing

UOS AI’s FollowAlong feature lets you select text and ask for a summary, translation or explanation. Saved selection actions can apply a prompt to the text you are reading.

1. Inspect the current deepin download and UOS AI documentation.
2. Check whether your chosen assistant uses an online service or a local model.
3. Choose a small reading or writing task to judge whether the selection workflow helps.

AnduinOS: familiar layout and optional local tools

AnduinOS offers an Ubuntu-based desktop with optional Why AI and offline dictation tools. Local execution depends on installing the required package and model.

1. Read the Why AI setup instructions linked below.
2. Check the model requirements against your available memory and hardware.
3. For voice typing, select the microphone and shortcuts and download the required recognition model before checking offline behaviour.

Linux-AI OS: distinguish the editions

The video describes the project’s advertised mail, image, voice and creative tools. Its free Star edition and paid suite have different scopes. Luke has not tested the paid suite.

1. Compare the current edition pages rather than assuming every advertised tool is included free.
2. Check GPU memory and other requirements for the actual image or voice tool you want.
3. Account for first-launch model downloads; local AI processing does not remove the mail service’s network connection.

Marlow: a developer preview of parallel input

Marlow’s demonstrations show a person working while an agent uses another window with separate input. The project describes connections to Gemini and Google services, rather than an all-local system.

1. Watch the developer’s own demonstrations from the official project site.
2. Distinguish the demonstrated behaviour from a public operating-system release: the video describes no public release to install.
3. Compare the parallel-input idea with the tasks where a computer-use agent currently interrupts your work.

The takeaway

Choose the workflow you want first: desktop customisation, help beside selected text, optional local tools, a packaged suite or a preview of parallel human-and-agent work.