Open-weight model
Muse-Glimmer-30B-Abliterated-MM-oQ5e
by Joey W Trampush, PhD brainworkup/Muse-Glimmer-30B-Abliterated-MM-oQ5e
Muse-Glimmer-30B-Abliterated-MM-oQ5e is an open-weight model from Joey W Trampush, PhD. It has 29.8B parameters and a 131,072-token context. At 16-bit it needs about 71.5 GB of GPU memory, which fits on 1x MI300X from $1.85 an hour, at the lowest prices in the SAVRN Index. It draws 93 downloads a month.
This model was quantized using oQ (oMLX v0.6.3rc1) mixed-precision quantization.
Runs On
What it takes to serve Muse-Glimmer-30B-Abliterated-MM-oQ5e (29.8B parameters): the memory its weights need at each precision, and the cheapest way to rent enough data-center GPUs to hold them.
| Precision | Weights | Memory needed | Cheapest setup | Per hour | Also fits |
|---|---|---|---|---|---|
| 16-bit | 59.6 GB | 71.5 GB | 1x MI300X (192 GB) Vultr |
$1.85 | 1x H100 $1.99 · 1x MI325X $2.00 |
| 8-bit | 29.8 GB | 35.7 GB | 1x MI300X (192 GB) Vultr |
$1.85 | 1x H100 $1.99 · 1x MI325X $2.00 |
| 4-bit | 14.9 GB | 17.9 GB | 1x MI300X (192 GB) Vultr |
$1.85 | 1x H100 $1.99 · 1x MI325X $2.00 |
Memory is the weights at that precision plus 20% for the runtime and a short context; a long context needs more. Prices are the lowest on-demand hourly rates in the SAVRN Index, read Oct 18, 2026.
Muse-Glimmer-30B-Abliterated-MM-oQ5e on every accelerator the SAVRN Index prices, at every precision
Model Card
This model was quantized using oQ (oMLX v0.6.3rc1) mixed-precision quantization.
Excerpt from the card by Joey W Trampush, PhD.
Configuration
- Architecture
- MuseGlimmerForConditionalGeneration
- Context length (tokens)
- 131,072
- Layers
- 52
- Hidden size
- 6,656
- Feed-forward size
- 19,968
- Attention heads
- 32
- Key/value heads
- 2
- Head dimension
- 128
- Vocabulary size
- 202,048
- Sliding window (tokens)
- 2,048
- Model type
- muse_glimmer
Identity and Version
- Repository
- brainworkup/Muse-Glimmer-30B-Abliterated-MM-oQ5e
- Publisher
- Joey W Trampush, PhD
- Task
- Not stated by the source
- Modality
- Other
- Library
- mlx
- Parameters
- 29.8B parameters
- Languages
- mlx, oq
- Revision
- bde3a71b70a7783f098a4b1a183d123a31017287
- First published
- 2026-08-20
- Last updated
- 2026-10-09
Files and Weights
15 files, 23.8 GB in total. The weights are 5 files totalling 23.7 GB in safetensors.
Every file
| File | Type | Size | SHA-256 |
|---|---|---|---|
| model-00001-of-00005.safetensors | Weights | 5.0 GB | 01a4dcd64bd3 |
| model-00002-of-00005.safetensors | Weights | 5.0 GB | 914423ac0758 |
| model-00003-of-00005.safetensors | Weights | 5.1 GB | 3fad8b6d5aca |
| model-00004-of-00005.safetensors | Weights | 5.0 GB | 4d15713f240e |
| model-00005-of-00005.safetensors | Weights | 3.7 GB | c48bf4ad6ecd |
| config.json | Configuration | 13.1 KB | — |
| generation_config.json | Configuration | 148 B | — |
| model.safetensors.index.json | Configuration | 208.7 KB | — |
| oq_imatrix_report.json | Configuration | 25.3 KB | — |
| processor_config.json | Configuration | 395 B | — |
| README.md | Documentation | 720 B | — |
| chat_template.jinja | Other | 7.3 KB | — |
| .gitattributes | Repository | 1.6 KB | — |
| tokenizer.json | Tokenizer | 28.1 MB | c9dbee66967b |
| tokenizer_config.json | Tokenizer | 80.0 KB | — |
License and Download
- License
- Not stated by the source
- Access
- Open weights, no gate
- Download size
- 23.7 GB
Released by Joey W Trampush, PhD through its official repository on Hugging Face.
Built From
- Derived from divinetribe/Muse-Glimmer-30B-Abliterated-MM-bf16
- Quantized from divinetribe/Muse-Glimmer-30B-Abliterated-MM-bf16
Memory Requirements
| Precision | Weights in memory |
|---|---|
| As published | 23.7 GB |
| 16-bit | 59.6 GB |
| 8-bit | 29.8 GB |
| 4-bit | 14.9 GB |
Weights only, from the published parameter count; the key-value cache and runtime add to this.
Questions About Muse-Glimmer-30B-Abliterated-MM-oQ5e
How much GPU memory does Muse-Glimmer-30B-Abliterated-MM-oQ5e need?
About 71.5 GB at 16-bit and 17.9 GB at 4-bit: the weights (29.8B parameters) plus a working margin. A long context needs more.
What is the cheapest GPU to run Muse-Glimmer-30B-Abliterated-MM-oQ5e on?
At 16-bit, 1x MI300X from $1.85 an hour; at 4-bit, 1x MI300X from $1.85 an hour, at the lowest on-demand prices the SAVRN Index lists.
What is Muse-Glimmer-30B-Abliterated-MM-oQ5e's context length?
131,072 tokens, from the maximum position embeddings in its published configuration.