AI Library
Model Information & Download
Model Options:
latest (18GB)
30b (18GB)
30b-bf16-dflash (59GB)
30b-mlx (19GB)
30b-mlx-bf16-dflash (65GB)
30b-mlx-bf16 (60GB)
30b-mxfp8 (33GB)
30b-mxfp8-dflash (35GB)
30b-nvfp4 (17GB)
30b-nvfp4-dflash (19GB)
30b-q4_K_M-dflash (20GB)
30b-q8_0-dflash (33GB)
30b-q4_K_M (18GB)
30b-q8_0 (31GB)
30b-bf16 (57GB)
Model Tag:
muse-glimmer
📋
Meta's latest open model built for always-on local agents. 30B parameters, licensed under Apache 2.0 and runs on a single GPU — tuned for tool use, long tasks, and failure recovery.
Model File Size
18 GB
Quantization
Q4
License
Not available
Last Updated
2026-09-04 (3 weeks ago)
Tutorial
Download and Run on your PC
← Back to Model Library