Runs on

Ollama

Local LLM server on the laptop — vendored, offline, no external account.

See features

Overview

PortableAI ships a vendored Ollama binary on the laptop. After setup it serves the model locally. There is no Ollama cloud account, and chat does not call an external API.

What that means

  • The server process runs on hardware you hold.
  • Pairing from the iOS app uses the local hotspot or LAN and a PIN.
  • You can load compatible models instead of a fixed catalog.

Not a SaaS integration

This is not a connection to someone else’s dashboard. It is the runtime PortableAI already runs.