Tech · October 15, 2026
Meta and Apple team up to make Muse AI models run better on Macs
Meta and Apple have announced a collaboration to optimize Meta's new Muse Spark Open AI model for Apple Silicon. The 816-billion-parameter model is designed to run locally on high-end Mac Studio machines using Apple's MLX framework, marking a significant shift in the historically contentious relationship between the two tech giants. While the model is open-weight and accessible via Hugging Face, running it requires substantial hardware investment, typically involving two to four Mac Studio units equipped with M5 Ultra chips and up to 1TB of unified memory.
Read the original coverage
- The VergeTrade
Meta and Apple team up to make Muse AI models run better on Macs
Read at The Verge ↗ - EngadgetTrade
Meta built a huge open-weight AI model for Macs, but you
Read at Engadget ↗ - AxiosMainstream
"Mommy and daddy" play nice: Apple and Meta team up to bring open-source Muse to Macs
Read at Axios ↗ - CNBCMainstream
Apple and Meta team up to make Muse AI model run on Mac computers
Read at CNBC ↗
Reported by multiple outlets
- Meta released the Muse Spark Open model, which features approximately 800-816 billion parameters.
- The model is optimized for Apple Silicon and the MLX framework.
- Meta CEO Mark Zuckerberg and new Apple CEO John Ternus publicly supported the launch.
- The model can run on two Mac Studio machines equipped with M5 Ultra chips.
- Muse Spark Open is available for download on Hugging Face.
Reported by only some outlets
- Running the model locally requires up to 1TB of RAM.
- Four Mac Studios with 256GB RAM each could run the model at a cost of roughly $38,000.
- Meta chief AI officer Alexandr Wang joined Meta following a $14 billion investment in his startup Scale AI.
- A single Mac Studio with 256GB RAM is priced at $10,799.
- Apple sued OpenAI in July alleging trade secret theft.
- Meta researcher Daniel Gross is cited as a key facilitator of the collaboration.
Industry-friendly coverage frames it as…
Focuses on the breakthrough of running massive models locally, the power of Apple's unified memory architecture, and the competitive positioning against Chinese AI labs.
Skeptical / consumer coverage frames it as…
Highlights the 'hefty investment' required, noting the hardware costs are 'squarely out of reach' for typical individual developers.
About these outlets
CadNews does not determine absolute truth or endorse political viewpoints. It compares publicly available reporting, identifies shared facts, highlights framing differences, and flags uncertainty.