Muse Glimmer 30B needs about 17.9 GB of memory at Q4_K_M (a 16.9 GB file) with an 8K context window, so it fits on a 24 GB GPU. This repository has 31 quants, from 4.8 GB (Q1_0) to 55.7 GB (BF16). Longer context needs more: at 32K tokens the KV cache alone adds about 1.7 GB.
Muse Glimmer 30B – a 30‑billion‑parameter multimodal agent model for local deployment
Muse Glimmer 30B is a 30‑billion‑parameter causal language model with a dedicated perception encoder. It supports text and image input, reliable tool use, multi‑step reasoning, and failure recovery. The model is optimized for local inference on consumer hardware.
Good at
- Multimodal text‑image reasoning
- Reliable tool‑based workflows
- Long‑horizon multi‑step planning
Limits
- Requires 8‑k context for memory estimates
- No external vision API needed
- Apache‑2.0 license limits commercial redistribution
Best for
- Building autonomous agents on local machines
- Debugging code in scaffolded environments
- Processing documents with embedded images
Who it’s for. Developers and researchers who need a local, multimodal agent capable of tool use and long‑horizon reasoning. The model is suitable for offline deployment on consumer GPUs with sufficient VRAM.
Quants & memory
| Quant | File | Est. memory | vs VRAM | Result | |
|---|---|---|---|---|---|
| Q1_0 | 4.8 GB | 5.8 GB | Fits | Get file → | |
| IQ1_S | 6.5 GB | 7.5 GB | Tight | Get file → | |
| IQ1_M | 7.1 GB | 8.0 GB | Tight | Get file → | |
| IQ2_XXS | 8.0 GB | 8.9 GB | Too big | Get file → | |
| IQ2_XS | 8.7 GB | 9.7 GB | Too big | Get file → | |
| Q2_0 | 9.0 GB | 9.9 GB | Too big | Get file → | |
| IQ2_S | 9.1 GB | 10.1 GB | Too big | Get file → | |
| IQ2_M | 9.8 GB | 10.8 GB | Too big | Get file → | |
| Q2_K_S | 10.0 GB | 11.0 GB | Too big | Get file → | |
| Q2_K | 10.7 GB | 11.6 GB | Too big | Get file → | |
| IQ3_XXS | 11.1 GB | 12.1 GB | Too big | Get file → | |
| IQ3_XS | 12.0 GB | 12.9 GB | Too big | Get file → | |
| Q3_K_S | 12.5 GB | 13.4 GB | Too big | Get file → | |
| IQ3_S | 12.5 GB | 13.5 GB | Too big | Get file → | |
| IQ3_M | 12.8 GB | 13.8 GB | Too big | Get file → | |
| Q3_K_M | 13.7 GB | 14.6 GB | Too big | Get file → | |
| Q3_K_L | 14.7 GB | 15.6 GB | Too big | Get file → | |
| IQ4_XS | 15.3 GB | 16.3 GB | Too big | Get file → | |
| Q4_0 | 16.0 GB | 17.0 GB | Too big | Get file → | |
| Q4_K_S | 16.1 GB | 17.1 GB | Too big | Get file → | |
| IQ4_NL | 16.1 GB | 17.1 GB | Too big | Get file → | |
| Q4_K_M | 16.9 GB | 17.9 GB | Too big | Get file → | |
| Q4_1 | 17.7 GB | 18.6 GB | Too big | Get file → | |
| Q5_0 | 19.3 GB | 20.3 GB | Too big | Get file → | |
| Q5_K_S | 19.3 GB | 20.3 GB | Too big | Get file → | |
| Q5_K_M | 19.8 GB | 20.7 GB | Too big | Get file → | |
| Q5_1 | 21.0 GB | 21.9 GB | Too big | Get file → | |
| Q6_K | 22.9 GB | 23.8 GB | Too big | Get file → | |
| Q8_0 | 29.6 GB | 30.5 GB | Too big | Get file → | |
| F16 | 55.7 GB | 56.7 GB | Too big | Get file → | |
| BF16 | 55.7 GB | 56.7 GB | Too big | Get file → |
For 8‑GB VRAM, use Q4_K_M (15.77 GB) or Q4_K_S (17.1 GB) – memory estimates are approximate. If 12‑GB VRAM is available, Q3_K_M (13.8 GB) or Q3_K_L (15.6 GB) provide higher quality. For 16‑GB VRAM, Q2_K (11.6 GB) or Q2_M (10.8 GB) are good choices.
What GPU do you need for Muse Glimmer 30B?
These figures add the file size, an fp16 KV cache at an 8K context window (worked out from this model’s own header) and about 0.5 GB for the runtime. They are estimates, so leave a little headroom, and expect to use less context or a smaller quant if you are right at the limit.
- 4 GB VRAM: no quant fits at an 8K context.
- 6 GB VRAM: the best quant that fits is Q1_0 (4.8 GB file, about 5.8 GB in use).
- 8 GB VRAM: the best quant that fits is IQ1_M (7.1 GB file, about 8.0 GB in use).
- 12 GB VRAM: the best quant that fits is Q2_K (10.7 GB file, about 11.6 GB in use).
- 16 GB VRAM: the best quant that fits is Q3_K_L (14.7 GB file, about 15.6 GB in use).
- 24 GB VRAM: the best quant that fits is Q6_K (22.9 GB file, about 23.8 GB in use).
- 32 GB VRAM: the best quant that fits is Q8_0 (29.6 GB file, about 30.5 GB in use).
- 48 GB VRAM: the best quant that fits is Q8_0 (29.6 GB file, about 30.5 GB in use).
Which Muse Glimmer 30B quant should you choose?
Q4_K_M is the usual starting point for Muse Glimmer 30B: it balances file size and quality well. Here it is a 16.9 GB file.
The Q8_0 file is 29.6 GB, about 1.7 times the size of Q4_K_M. It keeps the weights closer to the original, which only matters if you have the memory to spare.
If memory is tight, the smallest file is Q1_0 at 4.8 GB, with the quality trade-offs that come with fewer bits. Lower bit counts mean smaller files and more quality loss.
New to quant names? Read GGUF quantization explained.
Specifications
Architecture from the GGUF header
Source
Frequently asked questions about Muse Glimmer 30B
How much VRAM does Muse Glimmer 30B need?
About 17.9 GB at Q4_K_M with an 8K context window. That is the 16.9 GB file, plus the KV cache and roughly 0.5 GB of runtime overhead. Smaller quants need less and longer context needs more.
Can I run Muse Glimmer 30B on an 8 GB GPU?
Yes, with IQ1_M: the 7.1 GB file needs about 8.0 GB at 8K context. Larger quants in this repository do not fit in 8 GB.
What is the file size of Muse Glimmer 30B GGUF?
It depends on the quant: Q1_0 is 4.8 GB, IQ1_S is 6.5 GB, IQ1_M is 7.1 GB, IQ2_XXS is 8.0 GB, IQ2_XS is 8.7 GB, Q2_0 is 9.0 GB and 25 more.
What context length does Muse Glimmer 30B support?
The GGUF header lists a trained context length of 131,072 tokens (128K). The KV cache grows with the context you use: about 0.4 GB at 8K and 1.7 GB at 32K.
Where can I download Muse Glimmer 30B GGUF?
The files are on Hugging Face in the NANI-Nithin/Muse-Glimmer-30B-GGUF repository. The download buttons on this page open the file on Hugging Face. ggufmodels.com does not host any model files.
What license does Muse Glimmer 30B use?
Hugging Face lists the license as apache-2.0. Licenses can differ between the quantized files and the original model, so check the repository and the base model before commercial use.
Can I run Muse Glimmer 30B on 8 GB VRAM?
Use the Q4_K_M or Q4_K_S quant, which estimate 17–18 GB of VRAM at 8‑k context. This is close to the limit; performance may vary with batch size.
Does the model support vision?
Yes, it includes a dedicated perception encoder for interleaved text and image input.
What license applies to the model?
The model is released under the Apache‑2.0 license, allowing free use and modification.