MacPaw taps Liquid AI to offer on-device inference to devs building for its app store

Ukraine-based software developer MacPaw has announced a strategic partnership with Liquid AI to integrate locally hosted artificial intelligence into its product ecosystem. This collaboration aims to bolster MacPaw’s internal tools while eventually providing a robust tech stack for third-party developers building on its SetApp platform.

Enhancing Eney with On-Device Intelligence

The primary focus of this initiative is the evolution of Eney, MacPaw’s AI assistant introduced last year. By leveraging Liquid AI’s expertise, MacPaw is developing Elix, an on-device inference system paired with a local memory architecture. This shift toward local processing is designed to prioritize user privacy and security by ensuring data remains on the hardware rather than in the cloud.

"Before training our models, we select an architecture that is different and tailored to the hardware. That allows us to really have the most efficient version of intelligence that runs directly on the device, with benefits like privacy and security," said Ramin Hasani, co-founder and CEO of Liquid AI.

Empowering Offline Workflows

MacPaw CEO Oleksandr Kosovan emphasized that moving to locally hosted models will unlock new capabilities for users, specifically the ability to execute complex agentic workflows and AI-driven tasks while completely offline.

While Apple currently offers its own local model frameworks, Liquid AI differentiates itself by focusing on high-performance customization. Hasani noted that the company is building a "customization stack" that allows models to learn from user input, enabling them to become more intelligent and adaptable over time.

Scaling the SetApp Ecosystem

MacPaw is positioning its subscription-based app store, SetApp—which currently serves over 150,000 paying users—as a hub for AI-driven innovation. The company’s roadmap includes:

• Developer Access: Once the local processing architecture is finalized, MacPaw intends to open these tools to external developers, enabling them to integrate on-device inference into their own applications.
• Hybrid AI Hub: The platform will act as a centralized marketplace, offering access to both local models and cloud-based solutions from providers like Google.
• Credit-Based Pricing: MacPaw is currently testing a new billing model that charges users based on the complexity of AI operations and the number of tasks performed, ensuring a scalable cost structure for AI-heavy applications.

By bridging the gap between local privacy and cloud-based versatility, MacPaw is setting the stage for a more flexible, developer-friendly AI ecosystem.