The partnership will power MacPaw’s Eney assistant and later give SetApp developers access to local inference.

MacPaw will partner with Liquid AI to build locally hosted AI models for its products. The two companies will create an on-device inference system called Elix plus a local memory system. MacPaw plans to use this stack to power its Eney assistant with a locally hosted version.

MacPaw is a Ukraine-based app developer that operates the SetApp subscription store with over 150,000 paying users. Liquid AI builds models tailored to run directly on hardware. The company says its approach improves privacy and security by keeping processing on-device.

For developers, local inference means assistants and agentic workflows can run without a network connection. The SetApp store will also offer cloud models from Google and others as a one-stop shop. MacPaw is testing credit-based plans where users spend credits based on AI task complexity.

MacPaw expects to finish the local processing architecture with Liquid AI and then open the stack to third-party developers. Once available, app makers can use on-device inference inside SetApp. The company is also preparing credit-based pricing for AI operations in the store.

What matters

• MacPaw will use Liquid AI to build an on-device inference system called Elix for its apps.
• Builders can run assistants and agentic workflows offline when local models ship in SetApp.
• Watch for MacPaw to offer the local processing stack to SetApp developers after Elix lands.

Why it matters

Watch for MacPaw to offer the local processing stack to SetApp developers after Elix lands.

This GenAI News article was prepared in original wording using reporting and materials published by TechCrunch AI. Source reference: https://techcrunch.com/2026/08/05/macpaw-taps-liquid-ai-to-offer-on-device-inference-to-devs-building-for-its-app-store/.

Drafted by the GenAI News review pipeline.