Ukraine-based app developer MacPaw has partnered with Liquid AI to deploy locally hosted AI models that run on users' devices, and plans to make the technology available to external developers. The company is also preparing its subscription app store, SetApp, to support AI apps and is testing a credit-based pricing model.
Project components
MacPaw is currently developing an AI assistant called Eney, which it introduced last year. The company now aims to create a locally hosted version and has engaged Liquid AI to develop an on-device inference system named Elix and a local memory system to support it.
Ramin Hasani, co-founder and CEO of Liquid AI, told TechCrunch that they choose an architecture tailored to the hardware before training models. According to Hasani, this approach enables the most efficient intelligence to run directly on the device and brings advantages such as privacy and security.
Offline operation and customization
MacPaw CEO Oleksandr Kosovan said locally hosted models will let users run assistants and agentic workflows offline. Hasani added that Liquid AI’s models are optimized for performance across different capabilities, and the company is building a customization stack around the models so they can use user input and data to improve and become more adaptable over time.
Availability to developers and SetApp plans
After finalizing the local processing architecture with Liquid AI, MacPaw intends to offer the technology to developers so they can integrate on-device inference into their apps. Kosovan said the platform would also provide access to other cloud models from companies such as Google, effectively serving as a one-stop shop for developers.
SetApp, MacPaw’s subscription-based app store, currently has over 150,000 paying users. The company is experimenting with credit-based pricing on the store, where users can perform a number of AI operations determined by their credits and the complexity of the task.
Why this matters
The collaboration emphasizes local computation, which can enhance user privacy and security and enable offline functionality. At the same time, MacPaw aims to create a flexible ecosystem for developers by combining local and cloud model access and by making the on-device technology available beyond its own products.
