Instructions to use aj9o9/nvidia-nemotron-3.5-lightning-30b with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Notebooks
Google Colab
Kaggle
Local Apps Settings

How to use aj9o9/nvidia-nemotron-3.5-lightning-30b with llama.cpp:

Install (macOS, Linux)

curl -LsSf https://llama.app/install.sh | sh
# Start a local OpenAI-compatible server with a web UI:
llama serve -hf aj9o9/nvidia-nemotron-3.5-lightning-30b:Q4_K_M
# Run inference directly in the terminal:
llama cli -hf aj9o9/nvidia-nemotron-3.5-lightning-30b:Q4_K_M

Install from WinGet (Windows)

winget install llama.cpp
# Start a local OpenAI-compatible server with a web UI:
llama serve -hf aj9o9/nvidia-nemotron-3.5-lightning-30b:Q4_K_M
# Run inference directly in the terminal:
llama cli -hf aj9o9/nvidia-nemotron-3.5-lightning-30b:Q4_K_M

Use pre-built binary

# Download pre-built binary from:
# https://github.com/ggerganov/llama.cpp/releases
# Start a local OpenAI-compatible server with a web UI:
./llama-server -hf aj9o9/nvidia-nemotron-3.5-lightning-30b:Q4_K_M
# Run inference directly in the terminal:
./llama-cli -hf aj9o9/nvidia-nemotron-3.5-lightning-30b:Q4_K_M

Build from source code

git clone https://github.com/ggerganov/llama.cpp.git
cd llama.cpp
cmake -B build
cmake --build build -j --target llama-server llama-cli
# Start a local OpenAI-compatible server with a web UI:
./build/bin/llama-server -hf aj9o9/nvidia-nemotron-3.5-lightning-30b:Q4_K_M
# Run inference directly in the terminal:
./build/bin/llama-cli -hf aj9o9/nvidia-nemotron-3.5-lightning-30b:Q4_K_M

Use Docker

docker model run hf.co/aj9o9/nvidia-nemotron-3.5-lightning-30b:Q4_K_M

LM Studio
Jan

vLLM

How to use aj9o9/nvidia-nemotron-3.5-lightning-30b with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "aj9o9/nvidia-nemotron-3.5-lightning-30b"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "aj9o9/nvidia-nemotron-3.5-lightning-30b",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/aj9o9/nvidia-nemotron-3.5-lightning-30b:Q4_K_M

Ollama
How to use aj9o9/nvidia-nemotron-3.5-lightning-30b with Ollama:
```
ollama run hf.co/aj9o9/nvidia-nemotron-3.5-lightning-30b:Q4_K_M
```

Unsloth Studio

How to use aj9o9/nvidia-nemotron-3.5-lightning-30b with Unsloth Studio:

Install Unsloth Studio (macOS, Linux, WSL)

curl -fsSL https://unsloth.ai/install.sh | sh
# Run unsloth studio
unsloth studio -H 0.0.0.0 -p 8888
# Then open http://localhost:8888 in your browser
# Search for aj9o9/nvidia-nemotron-3.5-lightning-30b to start chatting

Install Unsloth Studio (Windows)

irm https://unsloth.ai/install.ps1 | iex
# Run unsloth studio
unsloth studio -H 0.0.0.0 -p 8888
# Then open http://localhost:8888 in your browser
# Search for aj9o9/nvidia-nemotron-3.5-lightning-30b to start chatting

Using HuggingFace Spaces for Unsloth

# No setup required
# Open https://huggingface.co/spaces/unsloth/studio in your browser
# Search for aj9o9/nvidia-nemotron-3.5-lightning-30b to start chatting

How to use aj9o9/nvidia-nemotron-3.5-lightning-30b with Pi:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama serve -hf aj9o9/nvidia-nemotron-3.5-lightning-30b:Q4_K_M

Configure the model in Pi

# Install Pi:
npm install -g @mariozechner/pi-coding-agent
# Add to ~/.pi/agent/models.json:
{
  "providers": {
    "llama-cpp": {
      "baseUrl": "http://localhost:8080/v1",
      "api": "openai-completions",
      "apiKey": "none",
      "models": [
        {
          "id": "aj9o9/nvidia-nemotron-3.5-lightning-30b:Q4_K_M"
        }
      ]
    }
  }
}

Run Pi

# Start Pi in your project directory:
pi

Docker Model Runner
How to use aj9o9/nvidia-nemotron-3.5-lightning-30b with Docker Model Runner:
```
docker model run hf.co/aj9o9/nvidia-nemotron-3.5-lightning-30b:Q4_K_M
```

Lemonade

How to use aj9o9/nvidia-nemotron-3.5-lightning-30b with Lemonade:

Pull the model

# Download Lemonade from https://lemonade-server.ai/
lemonade pull aj9o9/nvidia-nemotron-3.5-lightning-30b:Q4_K_M

Run and chat with the model

lemonade run user.nvidia-nemotron-3.5-lightning-30b-Q4_K_M

List all available models

lemonade list

Hermes Agent

How to use aj9o9/nvidia-nemotron-3.5-lightning-30b with Hermes Agent:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama serve -hf aj9o9/nvidia-nemotron-3.5-lightning-30b:Q4_K_M

Configure Hermes

# Install Hermes:
curl -fsSL https://hermes-agent.nousresearch.com/install.sh | bash
hermes setup
# Point Hermes at the local server:
hermes config set model.provider custom
hermes config set model.base_url http://127.0.0.1:8080/v1
hermes config set model.default aj9o9/nvidia-nemotron-3.5-lightning-30b:Q4_K_M

Run Hermes

hermes

Atomic Chat

OpenClaw

How to use aj9o9/nvidia-nemotron-3.5-lightning-30b with OpenClaw:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama serve -hf aj9o9/nvidia-nemotron-3.5-lightning-30b:Q4_K_M

Configure OpenClaw

# Install OpenClaw:
npm install -g openclaw@latest
# Register the local server and set it as the default model:
openclaw onboard --non-interactive --mode local \
  --auth-choice custom-api-key \
  --custom-base-url http://127.0.0.1:8080/v1 \
  --custom-model-id "aj9o9/nvidia-nemotron-3.5-lightning-30b:Q4_K_M" \
  --custom-provider-id llama-cpp \
  --custom-compatibility openai \
  --custom-text-input \
  --accept-risk \
  --skip-health

Run OpenClaw

openclaw agent --local --agent main --message "Hello from Hugging Face"

Nemotron-3.5-Lightning-30B-A3B GGUF

GGUF quants of nvidia/NVIDIA-Nemotron-3.5-Lightning-30B-A3B-BF16 for llama.cpp.

Built and tested on:

Hardware	Spec
GPU	NVIDIA GeForce RTX 3090 24 GB
System RAM	64 GB
Runtime	llama.cpp (recent master with `nemotron_h_moe` support)

Use a recent llama.cpp that includes Nemotron-H MoE (nemotron_h_moe). Older trees without that arch will not load these files.

Item	Value
Base model	`nvidia/NVIDIA-Nemotron-3.5-Lightning-30B-A3B-BF16`
GGUF arch	`nemotron_h_moe` (hybrid Mamba + Attention + MoE)
Params (approx.)	~30B total / ~3B active (128 experts, top-6)
Context (config)	up to 262144+ (train meta may report larger)
MTP head	included in GGUF (enable with `--spec-type draft-mtp`)
imatrix	not used
License	OpenMDW-1.1 (same family as base; see `LICENSE` if present, else base card)

Files

File	Quant	Size (approx.)
`NVIDIA-Nemotron-3.5-Lightning-30B-A3B-Q4_K_M.gguf`	Q4_K_M	~23.7 GiB
`NVIDIA-Nemotron-3.5-Lightning-30B-A3B-Q6_K.gguf`	Q6_K	~32.5 GiB

Benchmark (speed only)

Setup (this table): RTX 3090 24 GB + 64 GB RAM, llama.cpp llama-server, Q6_K,
-ngl 999 -ncmoe 22 -np 1 -fa on --jinja, large context (~200k slot).
No MTP / no speculative decoding for these numbers (--spec-type not used).
Measured with llama-benchy: pp4096 + tg256 at several context depths.

test	t/s	peak t/s
pp4096 @ d4096	690.52
tg256 @ d4096	60.65	74.00
pp4096 @ d8192	763.21
tg256 @ d8192	71.17	75.00
pp4096 @ d16384	763.03
tg256 @ d16384	72.98	75.00
pp4096 @ d32768	763.52
tg256 @ d32768	72.09	74.00
pp4096 @ d65536	750.57
tg256 @ d65536	71.03	73.00

Numbers are single-run; your results will vary with build flags, drivers, -ncmoe, and load.

Quick start

llama-server (no MTP) — matches the bench style

llama-server \
  -m NVIDIA-Nemotron-3.5-Lightning-30B-A3B-Q6_K.gguf \
  -ngl 999 \
  -ncmoe 22 \
  -np 1 \
  -fa on \
  --jinja \
  -c 200000 \
  --alias nemotron-q6k \
  --host 0.0.0.0 --port 8080

Q4_K_M: same flags, swap the -m path.

llama-server with MTP (speculative multi-token prediction)

MTP weights are in the GGUF. Enable built-in draft-MTP (no separate draft model file):

llama-server \
  -m NVIDIA-Nemotron-3.5-Lightning-30B-A3B-Q6_K.gguf \
  -ngl 999 \
  -ncmoe 22 \
  -np 1 \
  -fa on \
  --jinja \
  -c 200000 \
  --alias nemotron-q6k \
  --spec-type draft-mtp \
  --spec-draft-n-max 3 \
  --spec-draft-n-min 1 \
  --host 0.0.0.0 --port 8080

Tune --spec-draft-n-max (e.g. 2–4). Acceptance/speedup depends on workload; re-bench if you care about TG with MTP on.

llama-cli

llama-cli \
  -m NVIDIA-Nemotron-3.5-Lightning-30B-A3B-Q4_K_M.gguf \
  -ngl 999 \
  -ncmoe 16 \
  -c 8192 \
  -n 256

From this Hub repo

llama-server \
  --hf-repo aj9o9/nvidia-nemotron-3.5-lightning-30b \
  --hf-file NVIDIA-Nemotron-3.5-Lightning-30B-A3B-Q6_K.gguf \
  -ngl 999 -ncmoe 22 -np 1 -fa on --jinja -c 200000 \
  --alias nemotron-q6k

# with MTP
llama-server \
  --hf-repo aj9o9/nvidia-nemotron-3.5-lightning-30b \
  --hf-file NVIDIA-Nemotron-3.5-Lightning-30B-A3B-Q6_K.gguf \
  -ngl 999 -ncmoe 22 -np 1 -fa on --jinja -c 200000 \
  --alias nemotron-q6k \
  --spec-type draft-mtp \
  --spec-draft-n-max 3

VRAM / OOM (prefer expert offload, not lower `-ngl`)

Keep -ngl 999 (full GPU layer offload). If you OOM, offload MoE expert weights to system RAM:

Flag	Meaning
`-ncmoe N` / `--n-cpu-moe N`	MoE expert weights for the first N MoE layers on CPU/RAM
`-cmoe` / `--cpu-moe`	All MoE expert weights on CPU/RAM

# examples
llama-server -m ...-Q6_K.gguf -ngl 999 -ncmoe 8  -c 200000
llama-server -m ...-Q6_K.gguf -ngl 999 -ncmoe 22 -c 200000   # used for the speed table
llama-server -m ...-Q6_K.gguf -ngl 999 -cmoe      -c 200000   # max expert offload

Raise -ncmoe until it fits. Only lower -c if you still OOM after expert offload. 64 GB system RAM is what makes -ncmoe practical on a 3090-class box.

Prompting

ChatML-style template (<|im_start|> / <|im_end|>) with optional thinking and tool-calling. Prefer chat mode or /v1/chat/completions so the embedded template is applied. Server may suggest --reasoning-preserve for this template.

Quality notes

Straight K-quants from BF16 GGUF (no imatrix).
Q4_K_M: smaller / faster to load; Q6_K: higher fidelity (bench table above).
Small quality drop vs BF16/Q8; solid for chat and coding on 3090 + 64 GB RAM.

License

OpenMDW License Agreement, version 1.1 as with the base NVIDIA model materials.

Retain the license and applicable notices when redistributing.

Credits

Base model: NVIDIA
Runtime: ggml-org/llama.cpp

Disclaimer

Not affiliated with NVIDIA. Provided as-is. Follow the base model card for intended use, safety, and limitations.

Downloads last month: 140

GGUF

Model size

33B params

Architecture

nemotron_h_moe

Hardware compatibility

4-bit

6-bit

Model tree for aj9o9/nvidia-nemotron-3.5-lightning-30b

Base model

nvidia/NVIDIA-Nemotron-3.5-Lightning-30B-A3B-BF16

Quantized

(76)

this model

Collection including aj9o9/nvidia-nemotron-3.5-lightning-30b

Nemotron 3.5 lightning 30B

Collection

1 item • Updated 5 days ago

Nemotron-3.5-Lightning-30B-A3B GGUF

Files

Benchmark (speed only)

Quick start

llama-server (no MTP) — matches the bench style

llama-server with MTP (speculative multi-token prediction)

llama-cli

From this Hub repo

VRAM / OOM (prefer expert offload, not lower -ngl)

Prompting

Quality notes

License

Credits

Disclaimer

Model tree for aj9o9/nvidia-nemotron-3.5-lightning-30b

Collection including aj9o9/nvidia-nemotron-3.5-lightning-30b

VRAM / OOM (prefer expert offload, not lower `-ngl`)