Run Muse Glimmer 30B on M5 Ultra

Muse Glimmer 30B runs at 44 tok/s (43.7 as recorded) on a 256 GB M5 Ultra, a community run reported by omlx.ai (MLX, oQ5e, 1k context). It needs about 18 GB of unified memory. It is Meta's 30B dense model, released under the Apache 2.0 licence.

Speed
community44
tok/s
First Token
community0.78
seconds
RAM Needed
18
GB minimum
Engine
MLX
of this figure
Want to run Muse Glimmer 30B faster — or step up to bigger models? Find your Mac →

The figure

This is a community run: a real run, reported by its source and linked below. The index records it as given and never adjusts it.

MetricValue
Tokens per secondcommunity44 tok/s
Provenancecommunity run · omlx.ai
QuantizationoQ5e
EngineMLX
Time to first token0.78s
Prompt processing (prefill)community1663 tok/s at 8K context
Context1k
Figure dated2026-11-12
Minimum RAM18 GB
Parameters30B
Chip bandwidth1228 GB/s

Every figure for Muse Glimmer 30B on the M5 Ultra

Each row as the index holds it, with its source. “Mac RAM” is the memory of the Mac a run used, not the model’s requirement (18 GB). The headline figure follows the reference rule: the lower-median measured plain-decode run at 5 bits or fewer (a run whose quant was not stated counts only when no stated-quant run qualifies), else the estimate; a run marked “not used to anchor estimates” is listed here but never becomes the headline.

SpeedQuantEngineMac RAMContextSourceNote
community43.7 tok/soQ5eMLX256 GB1komlx.ai80-core GPU. oMLX v0.7.4.dev1, MTP/DFlash off; Muse Glimmer-oQ5e; 19.1 GB peak. 43.6 tok/s at 4k, 35.7 at 195k.
Provenance key
sourced
Published figure — vendor or press review
estimated
community
Linked individual run — blog, forum or submission

Speed by context depth: Muse Glimmer 30B on the M5 Ultra

Measured figures only, each from its source’s own session; the index does not estimate long-context or prefill speed, so a missing cell is —. Each @ column is plain decode with that many tokens already in the prompt, from the same session as the row’s short-context figure: decode slows as the KV cache it reads grows. Prefill is prompt processing, the tokens per second a runtime reads a prompt at; with long agent contexts it sets how long you wait before the first token.

RunDecode (row)@4K@8K@16K@32K@64K@128KPrefillTTFT
MLX · oQ5e · 256 GB Mac · omlx.aicommunity43.7 tok/s
1k
community43.6 tok/scommunity43 tok/scommunity42.4 tok/scommunity41.7 tok/scommunity40.1 tok/scommunity37.9 tok/scommunity1663 tok/s
at 8K
community0.78 s

Setup: run Muse Glimmer 30B on a Mac

Ollama — the index has not verified an Ollama tag for Muse Glimmer 30B, so it doesn’t print one (a guessed tag fails with “pull model manifest: file does not exist”). Search the Ollama library →

MLX — pick a build from Hugging Face’s MLX listings, then run it with its repo id:

pip install -U mlx-lm
mlx_lm.generate --model <repo-id> --prompt "Hello"

Downloads: Search the Ollama library → · Hugging Face model card →

The 44 tok/s figure above came from MLX at oQ5e; other runtimes and quants run at different speeds.

Prefer a GUI? LM Studio downloads GGUF and MLX builds in one click. Step by step: install Ollama on a Mac.

Muse Glimmer 30B on other Apple Silicon chips

ChipSpeedQuantEngineBandwidth
M5 Maxsourced27 tok/sQ4_K_MMLX614 GB/s
M4 Maxsourced24 tok/sQ4_K_MMLX546 GB/s
M4 Procommunity10 tok/sQ4_K_MLM Studio273 GB/s
M6community9 tok/squantization not statedLM Studio170 GB/s

About Muse Glimmer 30B

Meta's 30B model, ~27 tok/s on M5 Max. Apache 2.0 license, LLMCheck capability 42/50.

Apache 2.0 licence · 30B · 131K context · LLMCheck capability 42/50 · see its leaderboard rank →

Memory by context length

Estimated, not measured: the index’s minimum for Muse Glimmer 30B (18 GB — the weights at the quant that fits, plus ~2 GB) plus the KV cache for that many tokens, computed from its config.json. 13 full layers + 39 sliding-window layers (window 2048). A 16-bit KV cache is the default in llama.cpp, Ollama, LM Studio and mlx-lm; a q8 cache halves the KV part. Coding agents hold 64K tokens or more. Runtimes also add a buffer that grows with context, so leave a few GB spare. macOS lets the GPU use only part of a Mac’s memory by default (figures for an M5 Ultra Mac, which runs macOS 26 or later): “fits” means within the GPU memory a Mac of that tier can use with the limit raised, a tier tagged raise GPU limit fits only after sudo sysctl iogpu.wired_limit_mb raises it (the tooltip has the exact command), and the last column is the smallest Mac memory size whose default limit holds it (the GPU memory limit).

ContextKV cache (16-bit)RAM neededRAM, q8 KVSmallest tier that fitsWithin the default GPU limit
8K~0.2 GB~18.1 GB~18.0 GB24 GB Mac raise GPU limit32 GB Mac
32K~0.5 GB~18.4 GB~18.2 GB24 GB Mac raise GPU limit32 GB Mac
64K (coding-agent context)~1.0 GB~18.8 GB~18.4 GB24 GB Mac raise GPU limit32 GB Mac
128K~1.8 GB~19.7 GB~18.8 GB24 GB Mac raise GPU limit32 GB Mac

System requirements

Macs with the M5 Ultra

🛒 Get a Mac that runs Muse Glimmer 30B

Muse Glimmer 30B needs about 18 GB of unified memory. These Apple Silicon Macs have enough to run it. Apple builds the higher-memory configurations to order; Amazon does not stock them.

As an Amazon Associate, LLMCheck earns from qualifying purchases. These affiliate links cost you nothing extra and help keep the LLMCheck index free.

Compare More Models

See how Muse Glimmer 30B stacks up against other models on your specific Mac hardware.

Open Compare Tool Full Leaderboard