llama.cpp AI runtime with the tinyllama demo model, and an IPFS node. For a source-build walk-through, see macOS install (from source) — the steps are nearly identical on Linux.
Steps
1
Download
Grab the latest AppImage from Releases. The asset is named like
linux-x86_64-morpheus-app-<version>.AppImage (use linux-arm64-... on ARM64, e.g. Raspberry Pi). Mainnet builds have no suffix; testnet builds end in -test.2
Make executable and launch
Adjust the filename to the version you downloaded:
3
Let the app fetch its services
The first launch (and the first after each update) shows a Starting services screen: the app downloads the proxy-router, AI runtime, demo model, and IPFS node, then starts and probes each one. Container Runtime (Docker) is only detected, never installed — it is needed only for containerized agent features. Click Skip to continue without it; Retry re-runs failed steps.
4
Onboard
Accept terms, set a UI password, create or recover a wallet. Save the mnemonic somewhere safe.
5
Validate
Open
http://localhost:8082/swagger/index.html. Use Local Model in Chat to verify the pipeline.Cleanup
morpheus-app is the Electron package / userData folder name; the product name in the UI is MorpheusUI.)
See also: Troubleshooting.