Why Is Apple Betting on This Now?
Apple is betting that running AI will become a reason people choose a Mac. You can already see why in a demo Apple gave this year. A developer asked an AI running on his Mac to build a drawing app for the iPad. He tried it, decided he wanted rounder brushstrokes, and asked for the change. The Mac wrote the code and built the app again. Watch Apple’s WWDC demo.
That demo helps explain why Claris, an Apple company, is building Harbor. It also points to an interesting shift for anyone buying a computer. The next time you choose a Mac, one of your questions might be which AI models can I run on it?
Why local AI makes sense for Apple
When Apple introduced the M6 MacBook Pro in October, it specifically called out how quickly the laptop could process a request to a local AI model. Apple now gives buyers a reason to care about that speed alongside familiar tasks like editing photos or rendering video. Apple’s M6 MacBook Pro announcement.
But the hardware is only half the story. A better model can make a Mac you already own more useful. If a developer releases a model that does a better job of writing code and it fits on your machine, you can benefit from that progress with a download. Apple has something to gain even when another company builds the model.
What Harbor tells us about the next Mac
Harbor is designed to let an Apple silicon Mac run a downloadable AI model that other apps can use. You can start with one Mac. If a larger model needs more memory, Harbor can combine the memory of compatible Macs connected over Thunderbolt 5. As of November 12, 2026, Harbor is in private preview, and Claris has a signup page for updates. Claris Harbor preview.
Consider what that could mean for a small design studio. Today, one designer might run a model on a MacBook. In the near future, the studio could connect two office Macs to run a larger model and make it available to the team’s apps. The team could choose when to upgrade the model, when to add hardware, and which work to keep in-house.
Our read is that Apple sees an opportunity on both sides. Faster Macs give developers room to build more capable apps. Easier software gives people a reason to use the computing power they already paid for. Harbor is an early sign of where those two ideas could meet.
What you can do with local AI today
At Fenn, we use local AI to help you find things on your Mac. Say you remember a client approving a second print run, but you’ve forgotten the email subject and the filename. Type what you remember, and Fenn can find the saved message. The search runs on your Mac, your correspondence stays private, and you can keep working offline. Try Fenn for Mac.
Read also
Local processing • No files upload • Works offline