What are the hardware requirements for Muse Glimmer 30B QUASAR NVFP4 W4A4?

Muse Glimmer 30B QUASAR NVFP4 W4A4 requires approximately 23.5 GB of VRAM. Quantized variants may run on less VRAM; see the Quantized Models section on this page.

Who developed Muse Glimmer 30B QUASAR NVFP4 W4A4 and how large is it?

Muse Glimmer 30B QUASAR NVFP4 W4A4 is developed by QUASAR-QAT, a model with 30b parameters. The model is published as open weights on Hugging Face and indexed on LLM Explorer with full benchmark history.

Where can I download or evaluate Muse Glimmer 30B QUASAR NVFP4 W4A4?

Muse Glimmer 30B QUASAR NVFP4 W4A4 is hosted on Hugging Face and linked from this page. LLM Explorer also lists quantized variants and similar alternatives if available.

Muse Glimmer 30B QUASAR NVFP4 W4A4 by QUASAR-QAT — VRAM 23.5GB

Name: Muse Glimmer 30B QUASAR NVFP4 W4A4
Author: QUASAR-QAT

Muse Glimmer 30B QUASAR NVFP4 W4A4 is an open-source language model by QUASAR-QAT. Features: 30b LLM, VRAM: 23.5GB, License: apache-2.0, LLM Explorer Score: 0.3.

Arxiv:2608.13966 4-bit precision 4bit 8-bit Agentic Base model:meta-models/muse-gl... Base model:quantized:meta-mode... Blackwell Coding Compressed-tensors Conversational Endpoints compatible Fp4 Function-calling Image-text-to-text Long-context Multilingual Multimodal Muse-glimmer Muse glimmer Native-qat Nvfp4 Qat Quantization Quantization-aware-training Quantized Quasar Reasoning Region:us Safetensors Sharded Tensorflow Thinking Tool-calling Tool-use Vllm Vlm W4a4

Model Card on HF 🤗: https://huggingface.co/QUASAR-QAT/Muse-Glimmer-30B-QUASAR-NVFP4-W4A4

Muse Glimmer 30B QUASAR NVFP4 W4A4 Parameters and Internals

LLM Name	Muse Glimmer 30B QUASAR NVFP4 W4A4
Repository 🤗	https://huggingface.co/QUASAR-QAT/Muse-Glimmer-30B-QUASAR-NVFP4-W4A4
Base Model(s)	Muse Glimmer 30B meta-models/Muse-Glimmer-30B
Model Size	30b
Required VRAM	23.5 GB
Updated	2026-09-14
Maintainer	QUASAR-QAT
Model Type	muse_glimmer
Model Files	2.7 GB: 1-of-7 4.8 GB: 2-of-7 4.8 GB: 3-of-7 4.8 GB: 4-of-7 4.8 GB: 5-of-7 1.6 GB: 6-of-7 0.0 GB: 7-of-7
Model Architecture	MuseGlimmerForConditionalGeneration
License	apache-2.0
Model Max Length	131072
Transformers Version	5.16.1
Tokenizer Class	TokenizersBackend
Padding Token	<\|finetune_right_pad\|>

Rank the Muse Glimmer 30B QUASAR NVFP4 W4A4 capabilities

Have you tried this model? Rate its performance — this feedback helps the ML community find the right model for their needs.

Instruction Following and Task Automation
Factuality and Completeness of Knowledge
Censorship and Alignment
Data Analysis and Insight Generation
Text Generation
Text Summarization and Feature Extraction
Code Generation
Multi-Language Support and Translation

Best Alternatives to Muse Glimmer 30B QUASAR NVFP4 W4A4

Best Alternatives	Context / RAM	Downloads	Likes
Muse Glimmer 30B	0K / 59.6 GB	0	1052
Muse Glimmer 30B	0K / 59.6 GB	25869	12
Muse Glimmer 30B FP8 Block	0K / 34.4 GB	19413	9
Muse Glimmer 30B NVFP4	0K / 23.4 GB	7245	16
Muse Glimmer 30B NVFP4 W4A4	0K / 12.1 GB	10907	3
Muse Glimmer 30B QUASAR NVFP4	0K / 23.5 GB	75	2
Muse Glimmer 30B NVFP4	0K / 23.3 GB	1196	6
Muse Glimmer 30B Nvfp4	0K / 11.4 GB	1437	2
Muse Glimmer 30B JANG 6M	0K / 27.1 GB	915	2
...e Glimmer 30B MXFP8 Vision MLX	0K / 35.2 GB	912	1

Note: green Score (e.g. "73.2") means that the model is better than QUASAR-QAT/Muse-Glimmer-30B-QUASAR-NVFP4-W4A4.