Instructions to use QuietImpostor/Nemotron-3-Nano-REAP-21B-A3B-MXFP4-GGUF with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Notebooks
Google Colab
Kaggle
Local Apps Settings

How to use QuietImpostor/Nemotron-3-Nano-REAP-21B-A3B-MXFP4-GGUF 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 QuietImpostor/Nemotron-3-Nano-REAP-21B-A3B-MXFP4-GGUF:MXFP4
# Run inference directly in the terminal:
llama cli -hf QuietImpostor/Nemotron-3-Nano-REAP-21B-A3B-MXFP4-GGUF:MXFP4

Install from WinGet (Windows)

winget install llama.cpp
# Start a local OpenAI-compatible server with a web UI:
llama serve -hf QuietImpostor/Nemotron-3-Nano-REAP-21B-A3B-MXFP4-GGUF:MXFP4
# Run inference directly in the terminal:
llama cli -hf QuietImpostor/Nemotron-3-Nano-REAP-21B-A3B-MXFP4-GGUF:MXFP4

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 QuietImpostor/Nemotron-3-Nano-REAP-21B-A3B-MXFP4-GGUF:MXFP4
# Run inference directly in the terminal:
./llama-cli -hf QuietImpostor/Nemotron-3-Nano-REAP-21B-A3B-MXFP4-GGUF:MXFP4

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 QuietImpostor/Nemotron-3-Nano-REAP-21B-A3B-MXFP4-GGUF:MXFP4
# Run inference directly in the terminal:
./build/bin/llama-cli -hf QuietImpostor/Nemotron-3-Nano-REAP-21B-A3B-MXFP4-GGUF:MXFP4

Use Docker

docker model run hf.co/QuietImpostor/Nemotron-3-Nano-REAP-21B-A3B-MXFP4-GGUF:MXFP4

LM Studio
Jan
Ollama
How to use QuietImpostor/Nemotron-3-Nano-REAP-21B-A3B-MXFP4-GGUF with Ollama:
```
ollama run hf.co/QuietImpostor/Nemotron-3-Nano-REAP-21B-A3B-MXFP4-GGUF:MXFP4
```
Unsloth Desktop

How to use QuietImpostor/Nemotron-3-Nano-REAP-21B-A3B-MXFP4-GGUF with Pi:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama serve -hf QuietImpostor/Nemotron-3-Nano-REAP-21B-A3B-MXFP4-GGUF:MXFP4

Configure the model in Pi

# Install Pi:
npm install -g @earendil-works/pi-coding-agent
# Add to ~/.pi/agent/models.json:
{
  "providers": {
    "llama-cpp": {
      "baseUrl": "http://localhost:8080/v1",
      "api": "openai-completions",
      "apiKey": "none",
      "models": [
        {
          "id": "QuietImpostor/Nemotron-3-Nano-REAP-21B-A3B-MXFP4-GGUF:MXFP4"
        }
      ]
    }
  }
}

Run Pi

# Start Pi in your project directory:
pi

Docker Model Runner
How to use QuietImpostor/Nemotron-3-Nano-REAP-21B-A3B-MXFP4-GGUF with Docker Model Runner:
```
docker model run hf.co/QuietImpostor/Nemotron-3-Nano-REAP-21B-A3B-MXFP4-GGUF:MXFP4
```

Lemonade

How to use QuietImpostor/Nemotron-3-Nano-REAP-21B-A3B-MXFP4-GGUF with Lemonade:

Pull the model

# Download Lemonade from https://lemonade-server.ai/
lemonade pull QuietImpostor/Nemotron-3-Nano-REAP-21B-A3B-MXFP4-GGUF:MXFP4

Run and chat with the model

lemonade run user.Nemotron-3-Nano-REAP-21B-A3B-MXFP4-GGUF-MXFP4

List all available models

lemonade list

Hermes Agent

How to use QuietImpostor/Nemotron-3-Nano-REAP-21B-A3B-MXFP4-GGUF with Hermes Agent:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama serve -hf QuietImpostor/Nemotron-3-Nano-REAP-21B-A3B-MXFP4-GGUF:MXFP4

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 QuietImpostor/Nemotron-3-Nano-REAP-21B-A3B-MXFP4-GGUF:MXFP4

Run Hermes

hermes

Atomic Chat

OpenClaw

How to use QuietImpostor/Nemotron-3-Nano-REAP-21B-A3B-MXFP4-GGUF with OpenClaw:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama serve -hf QuietImpostor/Nemotron-3-Nano-REAP-21B-A3B-MXFP4-GGUF:MXFP4

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 "QuietImpostor/Nemotron-3-Nano-REAP-21B-A3B-MXFP4-GGUF:MXFP4" \
  --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"

Replacing REAP with REAM

by TomLucidor - opened Feb 12

Discussion

TomLucidor

Feb 12

Nice to see a small model like Nemotron-3-Nano made to be smaller and faster, but I wonder if there are ways for models to preserve as much quality as possible https://bknyaz.github.io/blog/2026/moe/
Bonus ask: Kimi-Linear, Nemotron-H series, GPT-OSS series, GLM-4.5-Air, Ring-V2/Ming-V2/Ling-V2 series, Seed-OSS, StepFun, and probably more deserves to be REAM-ed

QuietImpostor

Owner Feb 12

I was actually just reading about REAM! It does seem very intriguing, and I might try it with Nemotron. However, keep in mind I definitely won't be able to do all of the models you list, I was relying on Modal 's free credits to do just this one REAP. If you could possibly narrow it down, I could definitely try!

TomLucidor

Feb 12

•

edited Feb 13

Priority: Kimi-Linear (48B AND Linear), GPT-OSS-20B (good reasoner), GLM-4.7-Flash (30B), LongCat-Flash-Lite (see if embedding would cause issues), Nemotron-3-Nano (30B AND Linear), Ring-Mini-Linear-2.0 (16B AND Linear)
Optional: GPT-OSS-120B (OpenAI gift), GLM-4.5-AIr (106B and near-SOTA), DeepSeek-v2-lite (16B and legacy), Ring-Flash-Linear-2.0 (104B AND Linear)
Excessive: GLM-4.X/5, Qwen3-Coder, DeepSeek, Seed, Qwen3-235B, StepFun models, Ring-1T, Kimi-K2, MiniMax-M1, MiniMax-M2
Probably need to also note down Gemma, Granite-4.0-H, and whatever Mistral have on the list as well

QuietImpostor

Owner Feb 12

•

edited Feb 12

Some of those models are pretty big, but I'll see what I can do, no guarantees that I'll be able to do all of the priorities!

Also, remember it can't do dense models, so Llama 3.x and basic Mistral models won't work!

TomLucidor

Feb 12

Oh wait Llama 4 (the bad ones) were the ones with MoE, I forgot! Damn that was a long time ago. Mistral definitely has a lot of MoE models tho!
Also Kimi made Moonlight (16B) and I am surprised they made something on the consumer range!

QuietImpostor

Owner Feb 12

•

edited Feb 12

Yes, but the only Mistral MoE models are Mixtral (super old) and Mistral Large 3 (too large) Moonlight might work, but it's also quite old. Also, I just now realized you were the one who wrote the reddit post that taught me about REAM, what a small community! Anyways, I'm busy and won't be able to work on it anytime soon, so patience is appreciated.

Upload images, audio, and videos by dragging in the text input, pasting, or clicking here.

Tap or paste here to upload images

· Sign up or log in to comment