MacPaw is building an on-device AI layer that will work across your Mac apps

MacPaw has announced a long-term strategic partnership with Liquid AI to co-develop an on-device AI technology stack built specifically for the Mac. The work begins with Eney, MacPaw’s AI assistant for macOS, and is framed as an AI layer meant to operate across Mac apps rather than as a single-app feature.

The stack is designed to run on-device and is being built for the Mac platform, not ported from a generic cloud model pipeline. Starting inside Eney means the first surface is an existing macOS assistant, with the longer aim of a shared on-device layer that other Mac apps can use. No model sizes, latency targets, or hardware requirements were disclosed.

For engineers and builders on macOS, an on-device, Mac-specific AI layer matters because it points toward assistant behavior that can sit closer to local files, UI context, and app workflows without a mandatory round trip to a remote API. If the layer truly spans apps, it could become a common integration surface for Mac software that today wires AI in one product at a time.

In market terms, MacPaw is pairing product distribution on macOS with Liquid AI’s model and stack work instead of shipping only a thin client over third-party cloud models. That is a bet on platform-native, private, offline-capable AI as a differentiator on the Mac, where Apple’s own on-device features and a crowd of cloud copilots already compete for the same user attention.

What to watch next is how far the layer moves beyond Eney: whether third-party Mac apps get a clear way to plug into the same on-device stack, and whether MacPaw and Liquid AI publish concrete product milestones for that cross-app path. Until those details land, the partnership is a direction signal more than a finished platform.