Llama-3_1-Nemotron-Ultra-253B-v1-MLX-Q5 is an MLX Q5 checkpoint derived from nvidia/Llama-3_1-Nemotron-Ultra-253B-v1, intended for local text generation on Apple Silicon.

Intended use

Local text generation and chat-style prompting on Apple Silicon
MLX-LM experimentation with the declared upstream model family
Offline or operator-controlled inference workflows

Out of scope

Safety-critical decisions without domain expert review
Claims of benchmark superiority not backed by published evaluation data
Non-MLX runtime guarantees; this card documents the shipped HF checkpoint, not every possible serving stack

Training and conversion metadata

Parameter	Value
Repository	`LibraxisAI/Llama-3_1-Nemotron-Ultra-253B-v1-MLX-Q5`
Base model	`nvidia/Llama-3_1-Nemotron-Ultra-253B-v1`
Task	`text-generation`
Library	`mlx`
Format	MLX / Apple Silicon checkpoint
Quantization	Q5
Architecture	DeciLMForCausalLM
Model files	37
Config model_type	`nemotron-nas`

This card only reports metadata present in the Hugging Face repository, existing card frontmatter, or public config files. Missing benchmark, dataset, or training-run details are left explicit rather than reconstructed.

Tested inference path

Inference for this checkpoint has been tested with LibraxisAI/mlx-batch-server.
This is the recommended tested path for operator-controlled local inference on Apple Silicon.

Aspect	Status
Tested runtime	`LibraxisAI/mlx-batch-server`
Target hardware	Apple Silicon
Inference mode	Local / self-hosted
Hugging Face Hosted Inference	Disabled for this repository (`inference: false`)

This does not claim compatibility with every possible serving stack. It documents the path that has been exercised for this published checkpoint.

Usage

CLI

bash

pip install mlx-lm

mlx_lm.generate \
  --model LibraxisAI/Llama-3_1-Nemotron-Ultra-253B-v1-MLX-Q5 \
  --prompt "Summarize the key signals in this document and list the next action items." \
  --max-tokens 400

Python

python

from mlx_lm import load, generate

model, tokenizer = load("LibraxisAI/Llama-3_1-Nemotron-Ultra-253B-v1-MLX-Q5")

prompt = "Summarize the key signals in this document and list the next action items."
response = generate(model, tokenizer, prompt=prompt, max_tokens=400)
print(response)

Multi-turn with the chat template

This checkpoint follows the tokenizer/chat-template contract inherited from nvidia/Llama-3_1-Nemotron-Ultra-253B-v1 when the template is present in the repository:

python

from mlx_lm import load, generate

model, tokenizer = load("LibraxisAI/Llama-3_1-Nemotron-Ultra-253B-v1-MLX-Q5")

messages = [
    {"role": "user", "content": "Summarize the key signals in this document and list the next action items."},
]
prompt = tokenizer.apply_chat_template(messages, add_generation_prompt=True, tokenize=False)
response = generate(model, tokenizer, prompt=prompt, max_tokens=400)
print(response)

Example output

No public sample output is currently declared for this checkpoint.

Quantization notes

Aspect	Original/base checkpoint	This checkpoint
Lineage	`nvidia/Llama-3_1-Nemotron-Ultra-253B-v1`	`LibraxisAI/Llama-3_1-Nemotron-Ultra-253B-v1-MLX-Q5`
Runtime target	Upstream runtime format	MLX on Apple Silicon
Quantization	Base precision or upstream-declared format	Q5
Published quality delta	Not declared in public metadata	Not declared in public metadata

Limitations

No public benchmarks for this checkpoint are declared in the model metadata.
No public benchmark claims are made by this card unless listed in the frontmatter.
Validate outputs on your own domain data before relying on this checkpoint.
Memory use and speed depend heavily on the exact Apple Silicon generation, unified-memory size, and prompt length.

License

other. Check the upstream/base model license as well when a base model is declared.

Citation

bibtex

@misc{libraxisai-llama-3-1-nemotron-ultra-253b-v1-mlx-q5,
  title = {Llama-3_1-Nemotron-Ultra-253B-v1-MLX-Q5},
  author = {LibraxisAI},
  year = {2026},
  howpublished = {\url{https://huggingface.co/LibraxisAI/Llama-3_1-Nemotron-Ultra-253B-v1-MLX-Q5}},
  note = {MLX checkpoint published by LibraxisAI}
}