Meta Releases Muse Glimmer, a 30B-Parameter Open-Source AI Model for Local Execution

Meta's Superintelligence Labs has launched Muse Glimmer, a 30-billion-parameter AI model under the permissive Apache 2.0 license. Designed to run locally on consumer hardware like a single GPU or Mac, the model is optimized for agent work and coding. It can be quantized to under 20GB and features a 131,000-token context window with improved generation speed. This strategic move aims to commoditize AI capabilities and bolster Meta's core business by providing free, on-device intelligence.