Meta has introduced Muse Spark Open, an open‑weight artificial‑intelligence model designed to run on Apple’s Mac computers, according to TechCrunch. The announcement was made on Thursday, with Meta branding the system as “the most capable” open‑weight model available in the United States.
What the model is
According to TechCrunch, Muse Spark Open contains 816 billion parameters and supports a one‑million‑token context window. The model is described as natively multimodal, meaning it can process multiple data types such as text, code, and visual inputs without needing separate subsystems. Meta also positions the model for complex coding tasks and agentic labor, indicating a focus on software development and autonomous workflow assistance.
How it runs on Macs
TechCrunch reports that the model is built to run locally on Apple computers, but it requires substantial hardware resources. Engadget, cited by TechCrunch, notes that the model needs at least 1 terabyte of RAM to operate effectively. The company recommends the Mac Studio equipped with the M5 Ultra chip as the optimal platform, suggesting that the hardware’s performance characteristics align with the model’s demanding computational profile.
Integrated tools and use cases
According to TechCrunch, Meta has paired Muse Spark Open with Muse Code, an AI‑powered coding terminal. The combined solution was deployed in several agentic workflows, including a Mastodon client where the model performed “bug fixes, new features, [and] full build” tasks. The company also reports that the model was used to run DaVinci Resolve, a high‑end video‑editing application, where it was capable of reviewing and cropping footage.
Open‑weight versus closed models
TechCrunch highlights that open‑weight models—those whose weights are publicly released—offer industry advantages not found in closed, proprietary alternatives. According to the source, these benefits include greater cost efficiency, transparency, and flexibility for developers and enterprises. Meta is making Muse Spark Open available now through multiple API platforms, enabling broader access for developers who wish to integrate the model into their own applications.
Brief analysis (Opinion)
In my view, Meta’s decision to target the Mac Studio with a high‑resource, open‑weight model signals a strategic push to capture the professional creator market, where Apple hardware dominates. By coupling the model with tools like Muse Code and showcasing real‑world tasks—such as fixing a Mastodon client and editing video—the company demonstrates practical utility beyond academic benchmarks. However, the 1 TB RAM requirement may limit adoption to a niche of power users and enterprises that can afford top‑tier Mac hardware.
Conclusion
According to TechCrunch, Meta’s launch of Muse Spark Open represents a significant addition to the open‑weight AI ecosystem, combining massive scale, multimodal capabilities, and direct compatibility with Apple’s high‑performance Mac lineup. While the hardware demands are steep, the model’s reported performance in coding and video‑editing workflows suggests it could become a valuable asset for developers and content creators seeking locally run AI solutions.
Frequently Asked Questions
What is the size of Muse Spark Open’s parameter count?**
According to TechCrunch, the model contains 816 billion parameters. **Q: Which Apple device does Meta recommend for running the model?** A: TechCrunch reports that the model runs best on a Mac Studio equipped with the M5 Ultra chip. **Q: What are some of the tasks Muse Spark Open has been used for?** A: As reported by TechCrunch, the model has been used for bug fixes and new feature development in a Mastodon client, as well as reviewing and cropping footage in DaVinci Resolve.