MacPaw, the company behind CleanMyMac and the alternative macOS app marketplace Setapp, is integrating Liquid AI's models to run its AI assistant Eney locally on device. The practical implication: no cloud round-trips, no data leaving the machine, and lower latency for end users.
The move matters because on-device inference has historically been constrained by model size and hardware limits. Liquid AI's architecture — based on Liquid Neural Networks rather than standard transformers — is specifically designed to run efficiently on edge hardware, making it a credible fit for consumer Mac deployments.
For developers building apps within MacPaw's ecosystem, this opens a concrete path to shipping AI features without depending on third-party API calls or managing cloud infrastructure. MacPaw intends to surface the on-device inference capability as something third-party developers can actually build against.
From a privacy standpoint, local inference is increasingly a selling point for professional and enterprise users who are cautious about sending sensitive data to external servers. MacPaw's positioning here aligns product differentiation with a genuine user concern.
If you're a developer targeting the Setapp distribution channel, watch for API access to this inference layer — it could meaningfully cut the cost and complexity of adding AI features to Mac apps while keeping you on the right side of user privacy expectations.