Meta is open-sourcing Muse Spark with a new open-weight model called Muse Spark Open, a move that underscores the company's commitment to open AI models. The 800-billion parameter model has been optimized for Apple Silicon, but running it locally requires a staggering 1TB of RAM.
Meta, which has argued that the US should not let Chinese labs lead on open-source AI, calls Muse Spark Open "the most capable US open-weight model to date." The company's benchmarks suggest the model outperforms two leading Chinese models, GLM 5.3 and Kimi K3, on agentic tasks and visual reasoning. It is also the first model optimized to run locally on Apple's M5 Ultra-equipped Mac Studio machines.
Meta explains that Apple's unified memory architecture lets the M5 Ultra's integrated GPU use up to 512 GB per Mac Studio, so two machines can hold an entire model. "The result is two Mac Studios behave like one supercomputer with a terabyte of unified memory," the company says.
Apple has not yet announced specific pricing or a ship date for the 512GB Mac Studio, which is expected in "late October." Muse Spark Open can also run on four 256GB M5 Ultra Mac Studios, which will cost roughly $38,000. That puts the setup well out of reach for individual developers, though larger organizations that have been spending heavily on tokens may find a few Mac Studios a worthwhile investment, provided the model performs as well as Meta claims. Muse Spark Open is available to download now from Hugging Face.
Source: Engadget · Summarized by HeadlinesBriefing