What are the hardware requirements for Muse Glimmer 30B AWQ INT4?

Muse Glimmer 30B AWQ INT4 requires approximately 24.1 GB of VRAM. Quantized variants may run on less VRAM; see the Quantized Models section on this page.

Who developed Muse Glimmer 30B AWQ INT4 and how large is it?

Muse Glimmer 30B AWQ INT4 is developed by cyankiwi, a model with 30b parameters. The model is published as open weights on Hugging Face and indexed on LLM Explorer with full benchmark history.

How does Muse Glimmer 30B AWQ INT4 perform on standard benchmarks?

Muse Glimmer 30B AWQ INT4 has the following published scores: LMSYS ELO 1426. Compare against reference models on this page or on the LLM Explorer leaderboards.

Muse Glimmer 30B AWQ INT4 by cyankiwi — VRAM 24.1GB

Name: Muse Glimmer 30B AWQ INT4
Author: cyankiwi

Muse Glimmer 30B AWQ INT4 is an open-source language model by cyankiwi. Features: 30b LLM, VRAM: 24.1GB, License: apache-2.0, Quantized, LLM Explorer Score: 0.43, ELO: 1426.

Arxiv:2504.13181 Arxiv:2602.06036 Awq Base model:meta-models/muse-gl... Base model:quantized:meta-mode... Compressed-tensors Conversational Endpoints compatible Image-text-to-text Muse glimmer Quantized Region:us Safetensors Sharded Tensorflow

Model Card on HF 🤗: https://huggingface.co/cyankiwi/Muse-Glimmer-30B-AWQ-INT4

Muse Glimmer 30B AWQ INT4 Benchmarks

LMSys ELO 1426 vs 1272 (so35) 12.1%

^nn.n% — How the model compares to the reference models: Anthropic Sonnet 3.5 ("so35"), GPT-4o ("gpt4o") or GPT-4 ("gpt4").

What is the LLM Explorer Rank (Score)

Muse Glimmer 30B AWQ INT4 Parameters and Internals

LLM Name	Muse Glimmer 30B AWQ INT4
Repository 🤗	https://huggingface.co/cyankiwi/Muse-Glimmer-30B-AWQ-INT4
Base Model(s)	Muse Glimmer 30B meta-models/Muse-Glimmer-30B
Model Size	30b
Required VRAM	24.1 GB
Updated	2026-08-13
Maintainer	cyankiwi
Model Type	muse_glimmer
Model Files	2.7 GB: 1-of-5 5.4 GB: 2-of-5 5.3 GB: 3-of-5 5.4 GB: 4-of-5 5.3 GB: 5-of-5
AWQ Quantization	Yes
Quantization Type	awq
Model Architecture	MuseGlimmerForConditionalGeneration
License	apache-2.0
Model Max Length	131072
Transformers Version	5.16.0.dev0
Tokenizer Class	TokenizersBackend
Padding Token	<\|finetune_right_pad\|>

Rank the Muse Glimmer 30B AWQ INT4 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 AWQ INT4

Best Alternatives	Context / RAM	Downloads	Likes
Muse Glimmer 30B AWQ FP8	0K / 34.5 GB	45	1
...r 30B Abliterated MLX 4bit AWQ	0K / 19.4 GB	565	1
Muse Glimmer Q4 MLX	128K / 15.7 GB	5949	6
Muse Glimmer 30B 8bit	0K / 11.6 GB	2935	2
Muse Glimmer 30B OptiQ 4bit	0K / 19.9 GB	360	1
Muse Glimmer 30B MLX 8bit	0K / 33.4 GB	107	1
...use Glimmer 30B Hermes Agentic	0K / 59.6 GB	37	2
Muse Glimmer 30B 4bit	0K / 19.4 GB	0	11
...e Glimmer 30B Unsloth Bnb 4bit	0K / 22.2 GB	0	7
...immer 30B Abliterated MLX 6bit	0K / 26.4 GB	1010	1

Note: green Score (e.g. "73.2") means that the model is better than cyankiwi/Muse-Glimmer-30B-AWQ-INT4.