exo connects your devices into a local AI cluster and spreads a model across them, so a model too big for one Mac can run on several. Its Mac app runs in the menu bar and finds your other devices on the local network.
The Mac app needs Apple silicon and macOS Sequoia 15.1 or later; RDMA over Thunderbolt needs macOS Tahoe 26.2 or later. Your total memory decides which models fit.
exo is free and open source (Apache 2.0). Inference runs on your own devices and can work offline once models are downloaded.
brew install --cask exoOfficial downloads only. MLX never re-hosts an app, and never runs a command for you.
Ranked against the run models locally listings, from real public data only. How the score works.
exo is a product of Exo Labs; MLX isn’t affiliated with it. Something wrong? Write to support@mlx.app.