From: <huggingface.co/mlx-community> Newsgroups: alt.freeweights.nemotronSubject: NVIDIA-Nemotron-3.5-Lightning-30B-A3B-4bit Date: 2026-11-15Commit: 55ac8c8Signed: 2DC28CC7B26A283B ✓ Text generation model with 31.6B parameters. Safetensors weights, 17.8 GB in all. Published by mlx-community under OpenMDW-1.1.
31.6B params 17.8 GB NVIDIA-Nemotron-3.5-Lightning-30B-A3B-4bit safetensors custom terms Facts Params 31.6B Size 17.8 GB Formats safetensors Task text generation License OpenMDW-1.1 custom terms read terms
fw
# custom terms: read the license before you download
$ fw get mlx-community/NVIDIA-Nemotron-3.5-Lightning-30B-A3B-4bit
# manifest sha256 0768a5825cee… · 17.8 GB · 12 files
$ fw verify ./NVIDIA-Nemotron-3.5-Lightning-30B-A3B-4bit mlx-community/NVIDIA-Nemotron-3.5-Lightning-30B-A3B-4bitHosted (signed) mlx-community/NVIDIA-Nemotron-3.5-Lightning-30B-A3B-4bit Last check: 2026-11-15 00:21 UTC
Get help: custom fine-tune or private deploy. We review your setup before we quote.
Quantized of NVIDIA-Nemotron-3.5-Lightning-30B-A3B-BF16 .
Works with
Runs with Ranked by fit to this model (safetensors), then by GitHub stars.
Show all 24 Fine-tune with Show all 29
Overview
Size 17.8 GB
Weights 17.8 GB (safetensors)
Parameters 31.6B
Formats safetensors
Task text generation
License OpenMDW-1.1 (custom)
Files 12, each with a sha256
Publisher mlx-community
Added 2026-11-15
Section alt.freeweights.nemotron Lineage
Original release from the publisher.
Base model: nvidia/NVIDIA-Nemotron-3.5-Lightning-30B-A3B-BF16
Where these bytes come from
Publisher at the pinned commit →You
Sources are tried in order. A file counts only when its size and sha256 match the manifest. Sources are checked when you ask for them, not on a timer.
Sources
Badge Host Last check Result huggingface.co 2026-11-15 00:21 UTC passed
Model card From the publisher at commit
55ac8c8 . Shown as plain text with links; images open at the publisher.
This model mlx-community/NVIDIA-Nemotron-3.5-Lightning-30B-A3B-4bit was converted to MLX format from nvidia/NVIDIA-Nemotron-3.5-Lightning-30B-A3B-BF16 using mlx-lm version 0.31.3 (4-bit, group size 64).
NVIDIA-Nemotron-3.5-Lightning-30B-A3B is a hybrid Mamba-2 / attention Mixture-of-Experts model (~31B total parameters, ~3B active per token).
Use with mlx#
from mlx_lm import load, generate
model, tokenizer = load("mlx-community/NVIDIA-Nemotron-3.5-Lightning-30B-A3B-4bit")
prompt = "Hello, who are you?"
if tokenizer.chat_template is not None:
messages = [{"role": "user", "content": prompt}]
prompt = tokenizer.apply_chat_template(
messages, add_generation_prompt=True
)
response = generate(model, tokenizer, prompt=prompt, verbose=True) Files 12 files · 17.8 GB Weights 17.8 GB (safetensors). Sizes and sha256 from the signed manifest.
Hide all 12 files with their sha256 File Size sha256 .gitattributes 1.57 KB 34448b82c17d60fec9b65b1f093c115ddbaadc04beb1b0140b6bfed2e012a930 README.md 1.44 KB 31d67cc7ae807a7984eee77d35cd7fd507289cc2ea420447e72f12f3e20d2e89 chat_template.jinja 9.87 KB 58933db77d3099b4f78c55a38347a72e1ea05b97d6bd8f38775303dc0194e0a9 config.json 2.93 KB 2d4bdc543efd8ccd68fccbf1231b1d1216ab2a0482229ea5638d448127973b27 generation_config.json 210 B 0ed94823ae737b507aef6af01e837041b7aa2fb0306793a310cd201d89a31ff4 model-00001-of-00004.safetensors 5.12 GB 12e2ba8a7f0990b7e34ddb25cf4043fe39b105215fe61a0b7dc1447be6b247b3 model-00002-of-00004.safetensors 5.29 GB a095b11193a9d58e2c67c83bc2aff0eedde8f76b77ff8ee4781fe08f6c3f5363 model-00003-of-00004.safetensors 5.29 GB c55808e0cf33a6b957d34ab57caf4185c4aed49c6ef15c7c69683e7af7e892cc model-00004-of-00004.safetensors 2.07 GB 9c541daef892060fffb831afb9e20251ff18c492df7bacd0bfee4d8e43124d89 model.safetensors.index.json 65.4 KB 4c12ae32746441ec65e94d8cf312462ac10ab0d3c80513bba3d1c1ca3f2ed9ca tokenizer.json 17.1 MB 623c34567aebb18582765289fbe23d901c62704d6518d71866e0e58db892b5b7 tokenizer_config.json 430 B ec380f0a6c0c2e294eb2435f5ec521fb2121cfc970e6ebf023d06829714a4eaa
Versions Each time the pinned upstream commit or file list changed in this repo. Newest first.
Upstream commit Date Contents 55ac8c8926 pinned now 2026-11-15 00:21 UTC 12 files · 17.8 GB
Use it fw
# custom terms: read the license before you download
$ fw get mlx-community/NVIDIA-Nemotron-3.5-Lightning-30B-A3B-4bit
# manifest sha256 0768a5825cee… · 17.8 GB · 12 files
$ fw verify ./NVIDIA-Nemotron-3.5-Lightning-30B-A3B-4bit mlx-community/NVIDIA-Nemotron-3.5-Lightning-30B-A3B-4bitTerms that come with these files. OpenMDW 1.1: keep the agreement and the copyright notices with the files when you pass them on.
Custom license (custom). Read the license text first. Open this panel to show the commands once you accept its terms. fw.py Checks the signature and every sha256, tries each source in order, resumes.
$ python3 fw.py get https://freeweights.allternit.com/manifests/mlx-community/NVIDIA-Nemotron-3.5-Lightning-30B-A3B-4bit.json --dir ./NVIDIA-Nemotron-3.5-Lightning-30B-A3B-4bit
$ python3 fw.py verify ./NVIDIA-Nemotron-3.5-Lightning-30B-A3B-4bit ./mlx-community__NVIDIA-Nemotron-3.5-Lightning-30B-A3B-4bit.jsonhf The Hugging Face CLI with one setting. Same repo id.
$ HF_ENDPOINT =https://freeweights.allternit.com hf download mlx-community/NVIDIA-Nemotron-3.5-Lightning-30B-A3B-4bit --revision 55ac8c89261109b36c04371cd3f479a4594208c8 --local-dir ./NVIDIA-Nemotron-3.5-Lightning-30B-A3B-4bittransformers Load the weights in Python.
import os
os.environ["HF_ENDPOINT" ] = "https://freeweights.allternit.com"
from transformers import AutoModelForCausalLM, AutoTokenizer
tok = AutoTokenizer.from_pretrained("mlx-community/NVIDIA-Nemotron-3.5-Lightning-30B-A3B-4bit", revision="55ac8c89261109b36c04371cd3f479a4594208c8")
model = AutoModelForCausalLM.from_pretrained("mlx-community/NVIDIA-Nemotron-3.5-Lightning-30B-A3B-4bit", revision="55ac8c89261109b36c04371cd3f479a4594208c8", device_map="auto")vLLM Serve an OpenAI-compatible API.
$ HF_ENDPOINT =https://freeweights.allternit.com vllm serve mlx-community/NVIDIA-Nemotron-3.5-Lightning-30B-A3B-4bit --revision 55ac8c89261109b36c04371cd3f479a4594208c8License Name OpenMDW-1.1 Id OpenMDW-1.1 Class Custom terms Redistribution click-through Listing Hosted (signed) Terms OpenMDW 1.1: keep the agreement and the copyright notices with the files when you pass them on. Decided License as declared by the publisher. No person read it. 2026-11-15 Text in README.md Text sha256 31d67cc7ae807a7984eee77d35cd7fd507289cc2ea420447e72f12f3e20d2e89 Terms gate Yes. The commands stay folded until you open them after reading the license. Read the license text
Licenses are checked against the license table . A custom license passes only after a person reads it.
Provenance From: <huggingface.co/mlx-community> Newsgroups: alt.freeweights.nemotronSubject: NVIDIA-Nemotron-3.5-Lightning-30B-A3B-4bit Date: 2026-11-15Commit: 55ac8c8Signed: 2DC28CC7B26A283B ✓ How to verify
$ curl -O https://freeweights.allternit.com/manifests/mlx-community/NVIDIA-Nemotron-3.5-Lightning-30B-A3B-4bit.json
$ curl -O https://freeweights.allternit.com/manifests/mlx-community/NVIDIA-Nemotron-3.5-Lightning-30B-A3B-4bit.json.minisig
$ minisign -Vm NVIDIA-Nemotron-3.5-Lightning-30B-A3B-4bit.json -P RWQ7KGqyx4zCLXtUFOqvfv8L3g+8fpKJ1FsuxPyOYMF+4KufbUxKq1Oq
$ shasum -a 256 <file> # compare with the Files tab A manifest whose signature fails must be rejected. A file whose size or sha256 does not match must be discarded. Verify a download
Model tree No model in the catalog names this one as its parent.
8.2B params 16.5 GB a3-rl-laion_nemotron-gym-agent-calendar-80-8B safetensors permissive Facts Params 8.2B Size 16.5 GB Formats safetensors Task reinforcement learning License Apache-2.0 permissive flip From: laion <huggingface.co/laion> Subject: a3-rl-laion_nemotron-gym-agent-calendar-80-8B 15B params 30.0 GB Apriel-Nemotron-15b-Thinker safetensors permissive Facts Params 15B Size 30.0 GB Formats safetensors Task text generation License MIT permissive flip From: ServiceNow-AI <huggingface.co/ServiceNow-AI> Subject: Apriel-Nemotron-15b-Thinker nemotron 45.3 GB Apriel-Nemotron-15b-Thinker-GGUF gguf permissive Facts Params not stated Size 45.3 GB Formats gguf Task text generation License MIT permissive flip From: <huggingface.co/lmstudio-community> Subject: Apriel-Nemotron-15b-Thinker-GGUF 99M params 710 MB Nemotron-3-Diarization gguf safetensors custom terms Facts Params 99M Size 710 MB Formats gguf, safetensors Task voice activity detection License OpenMDW-1.1 custom terms read terms flip From: <huggingface.co/nvidia> Subject: Nemotron-3-Diarization 638M params 5.66 GB nemotron-3.5-asr-streaming-0.6b gguf safetensors custom terms Facts Params 638M Size 5.66 GB Formats gguf, safetensors Task automatic speech recognition License OpenMDW-1.1 custom terms read terms flip From: <huggingface.co/nvidia> Subject: nemotron-3.5-asr-streaming-0.6b nemotron 236 GB nvidia_OpenCodeReasoning-Nemotron-14B-GGUF gguf permissive Facts Params not stated Size 236 GB Formats gguf Task text generation License Apache-2.0 permissive flip From: bartowski <huggingface.co/bartowski> Subject: nvidia_OpenCodeReasoning-Nemotron-14B-GGUF
mlx-community <huggingface.co/mlx-community>Newsgroups: alt.freeweights.nemotronSubject: 
lmstudio-community 
nvidia