Meta's first open LLM release since Llama 4, released under Apache 2.0. It directly competes with Gemma 4 31B and Qwen 27B. Even though it is a dense model, which makes it comparatively slow compared to all the MoEs, we prefer this model over similarly sized Qwens.
Specs
Params30B
LicenseApache-2.0
Similarity · VAIL
VAIL Fingerprint
0154:019e:02c7:03b0:068e:09d8:146d:6060
Explore other models with behavioral similarity to Muse-Glimmer-30B.
Capability · Artificial Analysis
AA Index
17.5
Months Behind Frontier
17.5 mo
claude-3-7-sonnet-thinking
24 Feb 2025
Adoption · Hugging Face
RAM @ 30d
Relative Adoption Metric: 0.43×. Below benchmark. Measured at 30 days.
Hugging Face Downloads
278.3K
last 30d
933.9K
all time
HF Likes
2K
Relative Adoption Metric contextualizes downloads against the model's size bucket.
Related Models
