Friday, October 16, 2026

Apple and Meta Team Up: Meta’s Muse Spark Open AI Model Will Run Locally on Mac Studios

Two of Silicon Valley’s longest-running rivals are working together. Meta has released Muse Spark Open, a large open-weight AI model, and partnered with Apple so developers can run it locally on Mac Studio computers, CNBC reported on 15 October.

What is Muse Spark Open?

  • An 816-billion-parameter model using a mixture-of-experts architecture.
  • Open-weight: developers can download its parameters and run it on their own hardware, under a license that allows limited commercial use.
  • Available from sites such as Hugging Face, tuned to run faster on Apple silicon.

Meta describes it as the most capable US open-weight model released so far, according to its research blog.

How it runs on a Mac

Users can link two or four Mac Studios with the M5 Ultra chip to use the model for coding, video analysis and agent-style work. Apple published benchmarks and documentation for its MLX framework, which can split a model across several machines.

The price of running it at home

This is not cheap. CNBC notes that a Mac Studio with the required M5 Ultra chip and 256GB of memory starts at $10,799, and the model needs at least two of them (four at that memory size).

Why local AI matters

Running a model on your own machines instead of in the cloud keeps data in-house and can cut ongoing costs, the companies say. It also shows Apple’s chips competing in a space where Nvidia has dominated.

An unlikely partnership

Apple and Meta clashed for years, especially over Apple’s 2021 privacy changes. With John Ternus now leading Apple, the two are finding common ground. Meta AI chief Alexandr Wang told CNBC that Muse Spark Open is not Meta’s most capable model, but is meant to get developers building.

More in our Tech section.

Sources: CNBC, 15 October 2026; Meta AI Research.

Featured image: Apple Studio Display and Mac Studio with Peripherals by AzureSaturn, CC0, via Wikimedia Commons.

RELATED ARTICLES

LEAVE A REPLY

Please enter your comment!
Please enter your name here