Muse Glimmer 30B NVFP4 needs about 18.0 GB of memory at its main file (a 17.0 GB file) with an 8K context window, so it fits on a 24 GB GPU. This repository has 7 quants, from 15.2 GB to 19.6 GB. Longer context needs more: at 32K tokens the KV cache alone adds about 1.7 GB.
Muse Glimmer 30B NVFP4 – 27.9B parameter vision‑language model in GGUF
Muse Glimmer 30B NVFP4 is a 27.9B parameter model optimized with FP4 weight and activation quantization. It supports text and image input and outputs text. The model is released under the Apache‑2.0 license and is available in seven GGUF tiers.
Good at
- Large context 131072 tokens
- FP4 quantization reduces disk and GPU memory
- Text‑image generation
Limits
- Vision tower not included – treat as text‑only
- Requires gabrielcosi’s llama.cpp fork
- No official benchmark scores
Best for
- Multimodal inference on GPUs with limited VRAM
- Research on VLMs with FP4 precision
- Deploying VLMs in resource‑constrained environments
Who it’s for. Researchers and developers needing a large vision‑language model that fits into limited GPU memory. The model is suitable for experimentation and deployment on systems that support the required llama.cpp fork and FP4 quantization.
Quants & memory
| Quant | File | Est. memory | vs VRAM | Result | |
|---|---|---|---|---|---|
| — | 15.2 GB | 16.1 GB | Too big | Get file → | |
| — | 15.9 GB | 16.8 GB | Too big | Get file → | |
| — | 16.7 GB | 17.6 GB | Too big | Get file → | |
| — | 17.0 GB | 18.0 GB | Too big | Get file → | |
| — | 18.0 GB | 18.9 GB | Too big | Get file → | |
| — | 18.3 GB | 19.2 GB | Too big | Get file → | |
| — | 19.6 GB | 20.5 GB | Too big | Get file → |
For 8 GB VRAM, use the 15.19 GB GGUF file (est. 16.1 GB memory at 8 k context). For 12 GB VRAM, the 15.85 GB file (est. 16.8 GB) is suitable. Larger VRAM allows the 19.55 GB file (est. 20.5 GB).
What GPU do you need for Muse Glimmer 30B NVFP4?
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: no quant fits at an 8K context.
- 8 GB VRAM: no quant fits at an 8K context.
- 12 GB VRAM: no quant fits at an 8K context.
- 16 GB VRAM: no quant fits at an 8K context.
- 24 GB VRAM: the best quant that fits is the main file (19.6 GB file, about 20.5 GB in use).
Which Muse Glimmer 30B NVFP4 quant should you choose?
If memory is tight, the smallest file is the first one in the table at 15.2 GB, with the quality trade-offs that come with fewer bits.
New to quant names? Read GGUF quantization explained.
Specifications
Architecture from the GGUF header
Source
Frequently asked questions about Muse Glimmer 30B NVFP4
How much VRAM does Muse Glimmer 30B NVFP4 need?
About 18.0 GB at its main file with an 8K context window. That is the 17.0 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 NVFP4 on an 8 GB GPU?
Not at an 8K context. The smallest file needs about 16.1 GB, so you would need a larger GPU or CPU offloading.
What is the file size of Muse Glimmer 30B NVFP4 GGUF?
It depends on the quant: file is 15.2 GB, file is 15.9 GB, file is 16.7 GB, file is 17.0 GB, file is 18.0 GB, file is 18.3 GB and 1 more.
What context length does Muse Glimmer 30B NVFP4 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 NVFP4 GGUF?
The files are on Hugging Face in the esatapedico/Muse-Glimmer-30B-NVFP4-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 NVFP4 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 NVFP4 on 8 GB VRAM?
Yes, the 15.19 GB GGUF file requires approximately 16.1 GB of VRAM at 8 k context, which is within an 8 GB VRAM budget with some overhead.
Does the model support vision tasks?
The vision tower is not included in the GGUF files; treat the model as text‑only for now.
What runtime is needed?
Use gabrielcosi’s llama.cpp fork, commit 9ec0ea18 or later, to load the model correctly.