G4-Jamet-26B-A4B-MK-1 is an open-weight model from Nakra. It has 25.8B parameters and a 262,144-token context. At 16-bit it needs about 61.9 GB of GPU memory, which fits on 1x MI300X from $1.85 an hour, at the lowest prices in the SAVRN Index.
Runs On
What it takes to serve G4-Jamet-26B-A4B-MK-1 (25.8B 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 | 51.6 GB | 61.9 GB | 1x MI300X (192 GB) Vultr |
$1.85 | 1x H100 $1.99 · 1x MI325X $2.00 |
| 8-bit | 25.8 GB | 31.0 GB | 1x MI300X (192 GB) Vultr |
$1.85 | 1x H100 $1.99 · 1x MI325X $2.00 |
| 4-bit | 12.9 GB | 15.5 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 23, 2026.
G4-Jamet-26B-A4B-MK-1 on every accelerator the SAVRN Index prices, at every precision
Model Card
The publisher has not written a card for this model.
Configuration
- Architecture
- Gemma4ForConditionalGeneration
- Context length (tokens)
- 262,144
- Layers
- 30
- Hidden size
- 2,816
- Feed-forward size
- 2,112
- Attention heads
- 16
- Key/value heads
- 8
- Head dimension
- 256
- Vocabulary size
- 262,144
- Experts
- 128
- Sliding window (tokens)
- 1,024
- Model type
- gemma4
Identity and Version
- Repository
- Hastagaras/G4-Jamet-26B-A4B-MK-1
- Publisher
- Nakra
- Task
- Not stated by the source
- Modality
- Other
- Library
- Not stated by the source
- Parameters
- 25.8B parameters
- Languages
- Not stated by the source
- Revision
- c1ef78c3c102322aee56157b3f73b1c7c2c18eee
- First published
- 2026-10-01
- Last updated
- 2026-10-01
Files and Weights
19 files, 51.6 GB in total. The weights are 11 files totalling 51.6 GB in safetensors.
Every file
| File | Type | Size | SHA-256 |
|---|---|---|---|
| model-00001-of-00011.safetensors | Weights | 3.1 GB | 11704b2d424b |
| model-00002-of-00011.safetensors | Weights | 4.9 GB | 18768a9bb119 |
| model-00003-of-00011.safetensors | Weights | 4.9 GB | cf75acd3bf01 |
| model-00004-of-00011.safetensors | Weights | 4.9 GB | c0c80f5a372f |
| model-00005-of-00011.safetensors | Weights | 4.9 GB | d5b12d93929e |
| model-00006-of-00011.safetensors | Weights | 4.9 GB | 4da4e4a210a5 |
| model-00007-of-00011.safetensors | Weights | 4.9 GB | 61394b2d2ad2 |
| model-00008-of-00011.safetensors | Weights | 4.9 GB | 15a33db6a0ef |
| model-00009-of-00011.safetensors | Weights | 4.9 GB | b8794a7a145f |
| model-00010-of-00011.safetensors | Weights | 4.9 GB | 1514a2c76200 |
| model-00011-of-00011.safetensors | Weights | 4.4 GB | 6982c7b67aa7 |
| config.json | Configuration | 3.8 KB | — |
| generation_config.json | Configuration | 203 B | — |
| model.safetensors.index.json | Configuration | 103.2 KB | — |
| processor_config.json | Configuration | 1.7 KB | — |
| chat_template.jinja | Other | 18.7 KB | — |
| .gitattributes | Repository | 1.6 KB | — |
| tokenizer.json | Tokenizer | 32.2 MB | cc8d3a0ce364 |
| tokenizer_config.json | Tokenizer | 3.1 KB | — |
License and Download
- License
- Not stated by the source
- Access
- Open weights, no gate
- Download size
- 51.6 GB
Released by Nakra through its official repository on Hugging Face.
Memory Requirements
| Precision | Weights in memory |
|---|---|
| As published | 51.6 GB |
| 16-bit | 51.6 GB |
| 8-bit | 25.8 GB |
| 4-bit | 12.9 GB |
Weights only, from the published parameter count; the key-value cache and runtime add to this.
Built on This Model
- Quantized fromG4-Jamet-26B-A4B-MK-1-i1-GGUF
- Derived fromG4-Jamet-26B-A4B-MK-1-i1-GGUF
Questions About G4-Jamet-26B-A4B-MK-1
How much GPU memory does G4-Jamet-26B-A4B-MK-1 need?
About 61.9 GB at 16-bit and 15.5 GB at 4-bit: the weights (25.8B parameters) plus a working margin. A long context needs more.
What is the cheapest GPU to run G4-Jamet-26B-A4B-MK-1 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 G4-Jamet-26B-A4B-MK-1's context length?
262,144 tokens, from the maximum position embeddings in its published configuration.