estimate, except measured file sizes

Can I run Muse Glimmer 30B locally?

29.6B parameters. Parameter count is the published figure.

Licence:

Apache 2.0 source

Context:

131,072 tokens

Modality:

text + image in, text out

Released:

Aug 2026 source

What it is

Meta's agentic model, distilled from Muse Spark, built for tool use on consumer hardware. Dense (no experts in the config). The card says about 29.6B parameters, including the vision encoder, of which about 1.8B is a ViT-G/14 perception encoder. Text and image in, text out. Context 131,072. Licence Apache 2.0. The download is not gated. The card's Model Release Date is August 2026. The Hugging Face repo was created on 9 Aug 2026. Knowledge cutoff 4 Jan 2026 (model card). text_config: 52 layers, hidden size 6656, 32 attention heads, 2 KV heads, head dim 128, vocab 202,048. 13 layers are full attention and 39 are sliding attention with a 2048-token window, so the full-attention share is 13/52. KV in this table is this site's FP16 estimate on those 13 full layers plus the 2048-token window. Official GGUFs need llama.cpp b10353 or newer (PR #26841, merged 10 Aug 2026, https://github.com/ggml-org/llama.cpp/pull/26841). Measured GGUF files, decimal GB, Hugging Face listing 20 Oct 2026, vision projector and drafter excluded: KQuant-17GB-Q4_K_M 16.76 GB and KQuant-Dynamic-Q4_K_XL 19.65 GB. About 4.5 and 5.3 bits per weight, estimated from those file sizes and 29.6B parameters. BF16 safetensors are 59.55 GB (2 shards) on the base repo; that row is not a GGUF. The memory table does not include the vision projector or the drafter: +1.40 GB mmproj for image input; +1.63 GB optional DFlash drafter. The GGUF card's rough totals are about 17 GB, about 19 GB with vision, and about 20 GB with vision and the drafter for the 17 GB quant, and about 20 / 22 / 23 GB for the dynamic quant. The base card says Q4_K_M targets 24 GB VRAM and Dynamic targets 32 GB VRAM, with degradation 1.0% and 0.2%. That is Meta-reported guidance, not this site's fit score. Ollama: muse-glimmer (about 18 GB) and muse-glimmer:30b-mlx (about 19 GB), https://ollama.com/library/muse-glimmer. Tech report posted 19 Oct 2026: https://x.com/AIatMeta/status/2112265822274302066.

Meta-reported speed

Meta-reported (model card, not a measurement on this site). K-Quant-17GB plus quantized DFlash, batch 1, greedy: RTX 5090 (llama.cpp) 74.9 tok/s, 233.4 tok/s with DFlash (3.1×); M4 Max 23.7 tok/s, 37.8 tok/s with DFlash; M5 Max (ExecuTorch) 26.6 tok/s, 50.2 tok/s with DFlash. Source: https://huggingface.co/meta-models/Muse-Glimmer-30B

Parameter source · Architecture source · Model page

Memory per quant

KV is at 8,192 tokens. Overhead is 0.8 GB. Both are estimates. A measured file size is labelled. The last column is the cheapest machine that fits the total fully.

Cheapest rig that fits

KQuant-17GB-Q4_K_M, full: GeForce RTX 3090 24GB, €1,224 (used (median of sold listings))

KQuant-17GB-Q4_K_M, slow, partly in system RAM: Budget PC with RTX 3060 12GB, slow, partly in system RAM, €1,357 (estimate)

Prices moved: see Building a local LLM rig in the RAMageddon.

On a specific GPU

• Muse Glimmer 30B on RTX 3060 12GB
• Muse Glimmer 30B on RTX 4060 Ti 16GB
• Muse Glimmer 30B on RTX 3090
• Muse Glimmer 30B on RTX 4090
• Muse Glimmer 30B on RTX 5080
• Muse Glimmer 30B on RTX 5090
• Muse Glimmer 30B on RX 7900 XTX
• Muse Glimmer 30B on Dual RTX 3090
• Muse Glimmer 30B on Dual RTX 4090