Local AI Advisor

Can Mac mini M6 32 GB run Muse Glimmer (30B)?

32 GB unified · 170 GB/s memory bandwidth · Muse Glimmer (30B) has 29.6B parameters.

Short answer

Yes, acceptable for chat; borderline for coding questions.

Best quantization for chat: Q4_K_M (16 GB of weights) on LM Studio. Generation: 5–8 tok/s, leaving about 7.3 GB free for your other apps.

How it rates for each workload

Same model and machine, different jobs. Agents and long documents send far bigger prompts than chat, so they need much more speed and memory.

WorkloadVerdictQuantGenerationFirst replyContext
Chatwith Open WebUIAcceptableQ4_K_M5–8 tok/s≈3 s8KDetails
Coding questionswith ContinueBorderlineQ4_K_M5–8 tok/s≈20 s16KDetails
Coding in a repositorywith AiderBorderlineQ4_K_M4.9–8 tok/s≈65 s32KDetails
Agentic codingwith OpenCodeBorderlineQ4_K_M4.9–8 tok/s≈90 s32KDetails
Long documents (64K)with Open WebUIBorderlineQ4_K_M4.8–8 tok/s≈4 min64KDetails

Speeds are bandwidth-based estimates; no direct benchmark exists for this pair. “First reply” is the time to read the workload’s opening prompt. How we calculate

Different tool, context size or workload?

These pages use typical settings. The full check models your exact setup.