MacPaw, the Ukrainian app developer best known for its Mac utility software, has partnered with Liquid AI to bring locally hosted AI models to its products.

The company eventually plans to offer the underlying on-device inference stack to developers building for its SetApp subscription app store.

MacPaw is already working on a locally hosted version of its AI assistant Eney, unveiled last year, and has enlisted Liquid AI to develop an on-device inference system called Elix along with a local memory system.

Liquid AI‘s CEO Ramin Hasani said the company selects architectures tailored to hardware before training, aiming for an efficient version of intelligence that runs directly on the device with privacy and security benefits.

MacPaw CEO Oleksandr Kosovan added that local models would let users run assistants and agentic workflows offline.

While Apple already provides local models to developers, Hasani argued that Liquid AI‘s models focus on performance across different capabilities and include a customization stack so models can adapt and improve with user data.

MacPaw plans to emphasize AI apps for SetApp, which has over 150,000 paying users, and is experimenting with credit-based pricing where users spend credits per AI operation depending on task complexity.

Once the local processing architecture is locked in, MacPaw intends to make the tech available to developers so they can use on-device inference in their apps.

The platform will also offer access to cloud models from companies like Google, acting as a one-stop shop for developers.