Meta launches Muse Spark Open, an AI model that runs on Macs
Meta released Muse Spark Open, an 816-billion parameter open-weight AI model designed to run locally on Macs, including Mac Studio with M5 Ultra.
Stock photo for illustration only, not from the actual event
- Meta launched Muse Spark Open, an open-weight AI model with 816 billion parameters.
- Designed to run locally on Apple computers, performing best on Mac Studio with M5 Ultra.
- Features a 1-million token context window and native multimodal capabilities.
- Currently available to access through multiple API platforms.
Meta has announced the launch of what it describes as the most capable open-weight model in the United States, named Muse Spark Open, on October 15, 2026. The new model stands out for being specifically engineered to work seamlessly with Apple's Mac computers.
Leveraging a massive scale and an extensive context window, the model claims superior performance over other popular open-weight options according to evaluation results released by the company. It is natively multimodal and built to tackle complex coding tasks and agentic workflows.
Running the model locally requires substantial hardware power. According to reports, users will need at least 1 terabyte of RAM, and the model operates best on a Mac Studio equipped with the M5 Ultra chip. Despite the heavy hardware demands, it serves as a powerful tool for desktop tasks and development.
Stock photo for illustration only, not from the actual event
Meta paired the model with Muse Code, its AI coding terminal, deploying it into various agentic workflows such as bug fixes and feature development for a Mastodon client. Furthermore, the company utilized it alongside DaVinci Resolve, demonstrating the model's capability to review and crop high-end video footage.
Meta's release of a heavy-duty open-weight model tailored for local Mac hardware highlights the growing trend toward high-performance local AI computing. While the 1-terabyte memory requirement places it out of reach for everyday consumers, it opens significant doors for developers and enterprise environments prioritizing local data processing and flexibility.
Open-weight models provide distinct industry advantages over proprietary alternatives, offering cost efficiency, transparency, and architectural flexibility. Meta is making the new model available immediately across various API platforms.
Source: TechCrunch
Found something wrong in this article? Report an issue with this article
Comments
Leave a Comment