What happened
Meta announced Muse Spark Open, which it calls the "most capable US open-weight model to date" and the first in its class to run on a desktop using two Apple M5 Ultra Mac Studios with the latest MLX framework.
Why it matters
Meta says the natively multimodal model has 816 billion total parameters and excels at long-horizon tasks, complex software engineering, and multi-step reasoning, while Apple published a technical post on MLX and the model.
WHO IT HITSCompanies and developers who want to run large open-weight AI models locally on their own hardware, rather than through a cloud service, may now have a desktop option that Meta says works on two Apple M5 Ultra Mac Studios.
Summaries like this, in your inbox every morning.
Meta and Apple are usually rivals, but they worked together on this launch and publicly supported each other. Meta CEO Mark Zuckerberg said on X that the weights were releasing today and that it runs great on new Mac Studios thanks to work with @johnternus and Apple, while new Apple CEO John Ternus replied with congratulations and said it was amazing to run a model of Muse Spark Open's caliber and scale so efficiently right from your desk. Meta chief AI officer Alexandr Wang told Axios that the companies have been collaborating over several months, and Apple also published a new technical post about MLX, Muse Spark Open, and two other open models. The model itself is described by Meta as a natively multimodal model with 816 billion total parameters that excels at long-horizon tasks, complex software engineering, and multi-step reasoning. Meta's claim that it is the "most capable US open-weight model to date" is its own characterization, and the value of the Apple partnership appears to rest on letting a model of that size run on a desktop rather than a cloud setup.
Pick your industry and the AI tools you use, and get news related to your work every day.
Free · 30 seconds with Google · unsubscribe anytimeWhat is AIToday? →
Ask AI anything about this article. The AI reads this article, earlier AIToday articles, and Wikipedia, and cites its sources. Q&As are published on this page for other readers too.
Nvidia launched Nemotron 3.5 Super VL, an open 120-billion-parameter multimodal model, plus Bolt 74B-A9B for l…
IBM outbound CTO Tyler Lynch said teams should start from customer outcomes, not roles
Japan Mobility Show Bizweek 2026 showcased Physical AI applications, including Toyota's basketball-playing rob…
Meta released Muse Spark Open, an open-weight model with 816 billion parameters (42 billion active), up to one…
Google pitched Gemini 4 Argon as one model for robot control, simulation and training-data review, saying it s…
Nvidia released Nemotron 3.5 Super VL, an open multimodal model with 120 billion parameters for enterprise age…
