Meta releases Muse Glimmer 30B as an open local agent model
Meta released Muse Glimmer 30B weights under Apache 2.0 for local agent workflows, tool use, coding, multimodal reasoning, and function calling. Quantized variants are designed to run on consumer hardware, widening access to persistent agent capabilities without cloud infrastructure.
0 comments · 0 votes
Sign in to join the discussion →
No comments yet. Start the discussion.
Why it moved the index
A dated primary release makes an agent-oriented, multimodal 30B model downloadable under a permissive license and practical on consumer machines. Local execution, tool use, broad framework support, and irreversible weight diffusion materially expand deployment and misuse reach, while the model's smaller size and developer-reported same-class benchmarks keep the assessment below frontier-model launch magnitude.
Assessment history
-
R1
Toward 62 · confidence 90
Initial inclusion from Meta's dated primary launch page and open-weight release for the exact Muse Glimmer 30B version.
11 Aug 2026
Share this page
-
DoomBench assesses “Meta releases Muse Glimmer 30B as an open local agent model” as evidence moving toward doom, with magnitude 62 and confidence 90 out of 100 in the deployment reach category.
-
The DoomBench assessment of “Meta releases Muse Glimmer 30B as an open local agent model” is based on reporting from Meta AI Research and records the editorial rationale, source quality, attribution, and revision history.
-
DoomBench summarizes “Meta releases Muse Glimmer 30B as an open local agent model” as follows: Meta released Muse Glimmer 30B weights under Apache 2.0 for local agent workflows, tool use, coding, multimodal reasoning, and function...
https://www.doombench.com/news/meta-releases-muse-glimmer-30b-as-an-open-local-agent-model-2026-08-10