Meta Releases Muse Glimmer, a 30B Open Agentic Model, and Debuts Muse Code

Meta releases Muse Glimmer, a new 30B open agentic model that runs on 18GB of RAM, shipped under a permissive Apache 2.0 license. The release lands alongside Muse Code, a terminal-native, multi-agent CLI/TUI built on Meta’s Muse Spark model that positions itself as directly competitive with Anthropic’s Claude Code.

Muse Glimmer: an agentic model built to run locally

Muse Glimmer is a dense, multimodal model engineered for offline execution on consumer hardware rather than cloud infrastructure. Key points from Meta’s announcement:

• 30B parameters, runnable on 18GB of RAM — well within reach of a single consumer GPU, PC, or Mac
• Open weights under the Apache 2.0 license, allowing commercial use and modification
• Built for always-on local agent workflows: multi-step reasoning, reliable tool use, multimodal understanding, and failure recovery bundled into one model
• Positioned for use cases like local coding agents and LLM-as-judge evaluation without a network dependency

Muse Code: Meta’s answer to Claude Code

Muse Code is a terminal-based, multi-agent coding CLI/TUI powered by Meta’s Muse Spark model. Meta is pitching it as directly competitive with Claude Code, built around high-throughput, asynchronous background execution — fanning tasks out into isolated Git worktrees so background sub-agents can work on parallel features without disturbing the active working environment.

Why it matters

Between Muse Glimmer’s local-first, permissively licensed design and Muse Code’s push into the terminal-agent space, Meta is signaling a broader agentic-AI strategy: put a capable open model in developers’ hands that runs without cloud dependency, and pair it with tooling that competes head-on with the current generation of AI coding agents.

Disclaimer:

All data and information provided on this blog are for informational purposes only. All the image sources used are for reference only. The author makes no representations as to the accuracy, completeness, correctness, suitability, or validity of any information on this blog and will not be liable for any errors, omissions, or delays in this information or any losses, injuries, or damages arising from its display or use. This is a personal view and the opinions expressed here represent my own and not those of my employer or any other organization.