Meta returns to open models with Zuckerberg's plan to out-copy China and sell compute by auction
Meta launched Muse Glimmer, a 30B parameter open agent model from Superintelligence Labs. The compressed model requires under 20GB of memory, enabling deployment on consumer hardware.
Meta has introduced Muse Glimmer, marking the debut of its Superintelligence Labs division. The release focuses on open-weight models designed for efficiency, specifically targeting agent-based tasks.
The model boasts a 30 billion parameter count but is optimized for compression. Once compressed, it operates with less than 20 GB of memory, allowing execution on standard consumer devices rather than requiring specialized data center infrastructure.
This move aligns with Mark Zuckerberg's broader strategy to compete aggressively in the open-source AI sector. By lowering hardware barriers and emphasizing compute accessibility, Meta aims to accelerate adoption and influence the trajectory of general artificial intelligence development.
This page provides an editorial summary based on publicly available information. It is not a republished article. Use the source link below for the original report.