MacPaw, the software company behind popular Mac utilities, is partnering with Liquid AI to build an artificial intelligence (AI) stack that runs directly on your Mac instead of sending data to the cloud. The collaboration aims to create AI assistants that are faster, more private, and work even without an internet connection. Eney, MacPaw's new AI assistant for macOS, will be the first product to use this technology, with results expected later this year.

Why Does Running AI Locally on Your Mac Matter?

For years, AI has largely lived in the cloud. When you ask a chatbot a question or use an AI writing tool, your data travels to a distant server, gets processed, and comes back to you. That approach works, but it creates three problems: privacy concerns, slower responses, and expensive cloud bills. Local AI, also called on-device inference, solves all three by running the AI model directly on your device.

The economics are compelling. Analysts now estimate that inference, not training, accounts for the majority of AI compute costs in production. Data transfer fees alone can consume up to 70 percent of total spending for applications that handle large amounts of data, like video or sensor analysis. Running AI on-device eliminates those network costs entirely.

"After almost two decades of building software for the Mac, MacPaw is evolving its standalone products into a connected ecosystem, with AI as a core technology we build and own. We believe intelligence should live where people work: private by design, fast by default, and be able to reach the cloud when that's the better tool," said Oleksandr Kosovan, CEO and founder of MacPaw.

Oleksandr Kosovan, CEO and founder of MacPaw

How Will MacPaw and Liquid AI Build This Technology?

• Efficient Foundation Models: Liquid AI will develop and fine-tune its foundation models, which are large AI systems trained on broad knowledge, specifically for macOS tasks. These models are designed to be smaller and faster than typical cloud-based AI, so they can run smoothly on consumer hardware without draining battery or storage.
• On-Device Inference Engine: MacPaw's Elix technology handles the actual processing of AI tasks locally on Apple silicon chips. This keeps the heavy computational work on your Mac rather than sending requests to distant servers.
• Persistent Memory Layer: MacPaw's Mnemos technology lets the AI assistant remember context from previous conversations. Over time, the assistant becomes more useful because it understands your preferences, work patterns, and needs without forgetting what you told it yesterday.

This combination of on-device intelligence, persistent memory, and native macOS task execution has not existed in a single Mac product before. The partnership is designed as shared infrastructure, meaning the same technologies could eventually be made available to thousands of Mac developers through Setapp, MacPaw's software marketplace.

"We build efficient foundation models and the tools around them so that companies can bring intelligence onto the devices their customers already use. This partnership will bring efficient, private, on-device LFMs to millions of Mac users," said Ramin Hasani, Co-Founder and CEO of Liquid AI.

Ramin Hasani, Co-Founder and CEO of Liquid AI

What Does This Mean for the Broader AI Industry?

The MacPaw-Liquid AI partnership reflects a larger shift happening across technology. For years, the industry assumed all AI would run in the cloud. But as organizations scale AI from experiments to everyday operations, the costs become unsustainable. Enterprise AI cost overruns now hit 79 percent, with 80 to 85 percent of enterprises missing their AI infrastructure forecasts by more than 25 percent.

Companies are now reassessing where AI workloads should run. AMD, the chip manufacturer, recently emphasized that enterprise AI will be inherently distributed, with intelligence spanning AI-capable personal computers, edge devices, enterprise infrastructure, and cloud environments. Rather than assuming all AI belongs in the cloud, organizations are identifying which tasks benefit from local processing and which truly need cloud resources.

For high-frequency, latency-sensitive workloads like voice ordering or computer vision tasks, the cost difference between cloud and on-device inference is substantial. Some estimates put the per-inference cost of an equivalent on-device model at a small fraction of the cloud API cost once volume climbs past a modest threshold.

What Makes This Partnership Different?

Liquid AI's foundation models already power AI features for companies like Mercedes-Benz, Insilico Medicine, and Shopify. The company, spun out of MIT, specializes in building highly efficient AI systems designed for real-world environments where latency, privacy, and performance matter most. MacPaw brings two decades of deep expertise in building software specifically for Mac users, understanding their workflows and preferences in ways that generic AI assistants cannot.

The partnership also signals a shift in how AI companies think about their role. Rather than competing solely on model size or raw processing power, companies like Liquid AI are focusing on efficiency, meaning getting powerful AI results with smaller models that use less energy and storage. This efficiency is what makes on-device AI practical for millions of everyday users.

As enterprises and individual users alike grapple with AI costs and privacy concerns, partnerships like this one suggest the future of AI is not one-size-fits-all cloud services, but rather a mix of local and cloud processing tailored to specific tasks and constraints. For Mac users, that future is arriving later this year with Eney.