Artemis-31B-v1.2-NVFP4 is an open-weight model from Jim Tebo. It has 18.5B parameters and a 262,144-token context. At 16-bit it needs about 44.3 GB of GPU memory, which fits on 1x MI300X from $1.85 an hour, at the lowest prices in the SAVRN Index.
This is a calibrated NVIDIA NVFP4 conversion of TheDrummer/Artemis-31B-v1.2, a fine-tune of google/gemma-4-31B. It is provided as a compressed-tensors safetensors checkpoint for runtimes that support Gemma 4 and nvfp4-pack-quantized.
Runs On
What it takes to serve Artemis-31B-v1.2-NVFP4 (18.5B 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 | 36.9 GB | 44.3 GB | 1x MI300X (192 GB) Vultr |
$1.85 | 1x H100 $1.99 · 1x MI325X $2.00 |
| 8-bit | 18.5 GB | 22.2 GB | 1x MI300X (192 GB) Vultr |
$1.85 | 1x H100 $1.99 · 1x MI325X $2.00 |
| 4-bit | 9.2 GB | 11.1 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 24, 2026.
Artemis-31B-v1.2-NVFP4 on every accelerator the SAVRN Index prices, at every precision
Model Card
This is a calibrated NVIDIA NVFP4 conversion of TheDrummer/Artemis-31B-v1.2, a fine-tune of google/gemma-4-31B. It is provided as a compressed-tensors safetensors checkpoint for runtimes that support Gemma 4 and nvfp4-pack-quantized. The original fine-tune is credited to TheDrummer; this repository contains a precision conversion. The checkpoint has two safetensors shards totaling 20,446,828,448 bytes (19.04 GiB), plus its configuration, tokenizer, chat template, processor configuration, recipe, and conversion manifest. Download the whole repository to use this format. The two weight shards alone are insufficient. - The tokenizer and chat template match the source checkpoint by SHA-256.…
Excerpt from the card by Jim Tebo.
Configuration
- Architecture
- Gemma4ForConditionalGeneration
- Context length (tokens)
- 262,144
- Layers
- 60
- Hidden size
- 5,376
- Feed-forward size
- 21,504
- Attention heads
- 32
- Key/value heads
- 16
- Head dimension
- 256
- Vocabulary size
- 262,144
- Sliding window (tokens)
- 1,024
- Model type
- gemma4
- Quantization
- compressed-tensors
Identity and Version
- Repository
- FaustianDeal/Artemis-31B-v1.2-NVFP4
- Publisher
- Jim Tebo
- Task
- Not stated by the source
- Modality
- Other
- Library
- Not stated by the source
- Parameters
- 18.5B parameters
- Languages
- Not stated by the source
- Revision
- 0d3668f9624fa59881e23cebabaf44b5239a94fd
- First published
- 2026-09-26
- Last updated
- 2026-09-26
Files and Weights
13 files, 20.5 GB in total. The weights are 2 files totalling 20.4 GB in safetensors.
Every file
| File | Type | Size | SHA-256 |
|---|---|---|---|
| model-00001-of-00002.safetensors | Weights | 20.0 GB | dbe043d04d5f |
| model-00002-of-00002.safetensors | Weights | 450.1 MB | 08ec9c8929bc |
| config.json | Configuration | 19.0 KB | — |
| conversion_manifest.json | Configuration | 450 B | — |
| generation_config.json | Configuration | 204 B | — |
| model.safetensors.index.json | Configuration | 254.2 KB | — |
| processor_config.json | Configuration | 1.7 KB | — |
| recipe.yaml | Configuration | 257 B | — |
| README.md | Documentation | 2.7 KB | — |
| chat_template.jinja | Other | 18.7 KB | — |
| .gitattributes | Repository | 1.6 KB | — |
| tokenizer.json | Tokenizer | 32.2 MB | cc8d3a0ce364 |
| tokenizer_config.json | Tokenizer | 3.7 KB | — |
License and Download
- License
- Not stated by the source
- Access
- Open weights, no gate
- Download size
- 20.4 GB
Released by Jim Tebo through its official repository on Hugging Face.
Built From
- Derived from TheDrummer/Artemis-31B-v1.2
- Quantized from TheDrummer/Artemis-31B-v1.2
Memory Requirements
| Precision | Weights in memory |
|---|---|
| As published | 20.4 GB |
| 16-bit | 36.9 GB |
| 8-bit | 18.5 GB |
| 4-bit | 9.2 GB |
Weights only, from the published parameter count; the key-value cache and runtime add to this.
Questions About Artemis-31B-v1.2-NVFP4
How much GPU memory does Artemis-31B-v1.2-NVFP4 need?
About 44.3 GB at 16-bit and 11.1 GB at 4-bit: the weights (18.5B parameters) plus a working margin. A long context needs more.
What is the cheapest GPU to run Artemis-31B-v1.2-NVFP4 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 Artemis-31B-v1.2-NVFP4's context length?
262,144 tokens, from the maximum position embeddings in its published configuration.