TPSTokens Per Second

Muse-Glimmer-30B: hardware requirements

Muse-Glimmer-30B needs about 20 GB of memory at Q4 quantization — a 18 GB download plus room for context and overhead. That fits a machine with 32GB RAM.

Parameters
30B
Tier
Needs a good desktop
Licence
See model card
Released
2026-08-10

Will it run on your machine?

8GB RAMNot enough memory16GB RAMNot enough memory24GB RAMTight fit32GB RAMRuns well64GB RAMRuns well128GB RAMRuns well8GB VRAMNot enough memory12GB VRAMNot enough memory16GB VRAMNot enough memory24GB VRAMRuns well32GB VRAMRuns well

Download sizes by quantization

Lower quantization means a smaller file and less memory, at some cost to quality. Q4_K_M is the usual starting point. What is this?

QuantDownloadRAM @ 4KRAM @ 32KSource
Q4_K_Mrecommended18 GB20 GB22 GBestimated
Q5_K_M19 GB21 GB23 GBmeasured
Q6_K24 GB27 GB29 GBestimated
Q8_032 GB35 GB37 GBmeasured
BF1660 GB63 GB65 GBmeasured

“Measured” sizes are read from published GGUF files (unsloth/Muse-Glimmer-30B-GGUF). “Estimated” sizes are derived from the parameter count and are typically within a few percent. The RAM columns add the context cache and runtime overhead to the download size.

Model card on HuggingFace →API specs & pricing →Verified 2026-09-14Listed for 120 days after release