Desktop vs Core: Which Do I Use
Wayland comes in two halves that share one engine. The desktop app is the graphical surface most people use day to day. Wayland Core is the same engine as a command-line tool, for scripting and servers. They read the same config and run the same agent loop, so picking one is about the interface, not about features that only one side has.
Reach for the desktop app when
Section titled “Reach for the desktop app when”- You want a graphical interface for chat, teams, and settings.
- You are connecting messaging channels like Telegram, Slack, or WhatsApp. Channel setup, pairing, and routing all live in the app.
- You are running multi-agent teams with a leader and specialist tabs.
- You want the scheduler UI, voice input and output, or image generation.
- You want to reach your agent from a phone or browser through remote access.
Reach for Wayland Core when
Section titled “Reach for Wayland Core when”- You are scripting Wayland or running it in CI.
- You are driving the engine from another program over the JSON stream protocol.
- You are running on a headless server with no display.
- You want a one-shot prompt in a shell pipeline.
What they share
Section titled “What they share”Both sides use the same engine, the same eight built-in tools, the same MCP client, and the same provider catalog (the engine runs five providers natively; the rest connect through the desktop). Config you write for one applies to the other. You can start a task in the app and continue the same kind of work from the CLI.
If you are not sure, start with the desktop app. Add the CLI later when you have a job that calls for automation. See Install Wayland Core (CLI).