Instructions to use Ninnix96/Ling-3.0-tiny-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 Ninnix96/Ling-3.0-tiny-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 Ninnix96/Ling-3.0-tiny-gguf:MXFP4
# Run inference directly in the terminal:
llama cli -hf Ninnix96/Ling-3.0-tiny-gguf:MXFP4

Install from WinGet (Windows)

winget install llama.cpp
# Start a local OpenAI-compatible server with a web UI:
llama serve -hf Ninnix96/Ling-3.0-tiny-gguf:MXFP4
# Run inference directly in the terminal:
llama cli -hf Ninnix96/Ling-3.0-tiny-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 Ninnix96/Ling-3.0-tiny-gguf:MXFP4
# Run inference directly in the terminal:
./llama-cli -hf Ninnix96/Ling-3.0-tiny-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 Ninnix96/Ling-3.0-tiny-gguf:MXFP4
# Run inference directly in the terminal:
./build/bin/llama-cli -hf Ninnix96/Ling-3.0-tiny-gguf:MXFP4

Use Docker

docker model run hf.co/Ninnix96/Ling-3.0-tiny-gguf:MXFP4

LM Studio
Jan

vLLM

How to use Ninnix96/Ling-3.0-tiny-gguf with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "Ninnix96/Ling-3.0-tiny-gguf"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "Ninnix96/Ling-3.0-tiny-gguf",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/Ninnix96/Ling-3.0-tiny-gguf:MXFP4

Ollama
How to use Ninnix96/Ling-3.0-tiny-gguf with Ollama:
```
ollama run hf.co/Ninnix96/Ling-3.0-tiny-gguf:MXFP4
```
Unsloth Desktop

How to use Ninnix96/Ling-3.0-tiny-gguf with Pi:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama serve -hf Ninnix96/Ling-3.0-tiny-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": "Ninnix96/Ling-3.0-tiny-gguf:MXFP4"
        }
      ]
    }
  }
}

Run Pi

# Start Pi in your project directory:
pi

Docker Model Runner
How to use Ninnix96/Ling-3.0-tiny-gguf with Docker Model Runner:
```
docker model run hf.co/Ninnix96/Ling-3.0-tiny-gguf:MXFP4
```

Lemonade

How to use Ninnix96/Ling-3.0-tiny-gguf with Lemonade:

Pull the model

# Download Lemonade from https://lemonade-server.ai/
lemonade pull Ninnix96/Ling-3.0-tiny-gguf:MXFP4

Run and chat with the model

lemonade run user.Ling-3.0-tiny-gguf-MXFP4

List all available models

lemonade list

Hermes Agent

How to use Ninnix96/Ling-3.0-tiny-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 Ninnix96/Ling-3.0-tiny-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 Ninnix96/Ling-3.0-tiny-gguf:MXFP4

Run Hermes

hermes

Atomic Chat

OpenClaw

How to use Ninnix96/Ling-3.0-tiny-gguf with OpenClaw:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama serve -hf Ninnix96/Ling-3.0-tiny-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 "Ninnix96/Ling-3.0-tiny-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"

Ling-3.0-tiny — Q4_K / MXFP4 / Q8_0 GGUF

Fast GGUF for mainstream llama.cpp. Works out of the box on any build that ships MXFP4 support (llama.cpp ≥ b3900).

Recipe (antirez style)

Tensor group	Format
Routed expert gate / up	Q4_K (calibrated)
Routed expert down	MXFP4
Shared experts, dense FFN (layer 0), attention, KDA, MLA Q-LoRA, output	Q8_0
Token embeddings	BF16
Routers, norms, SSM, gate biases	F32

Same quantization strategy as Salvatore Sanfilippo's Qwen3.8 Flash Next Q4: calibrated Q4_K on the large expert projections, MXFP4 on the down path, Q8_0 everywhere the signal matters most.

Files

File	Size
Ling-3.0-tiny-Q4K-MXFP4-Q8.gguf	4.74 GiB

Speed (llama.cpp)

Measured on a 12th-gen Intel laptop (4P+8E cores, 32 GB RAM).

Backend	Threads	Prefill pp512	Decode tg128
CPU	6	~57 tok/s	~20 tok/s
Vulkan (Intel Iris Xe)	2	~247 tok/s	~28 tok/s

Fast enough for real-time chat on CPU alone, no discrete GPU needed.

Usage

# interactive chat
llama-cli -m Ling-3.0-tiny-Q4K-MXFP4-Q8.gguf -t 4

# server
llama-server -m Ling-3.0-tiny-Q4K-MXFP4-Q8.gguf -t 4 --host 0.0.0.0 --port 8080

Model

Ling-3.0-tiny is a 128-expert MoE with 1.3 B active parameters out of 7.9 B total. Architecture: BailingMoE3 with KDA attention, MLA Q-LoRA compression, shared experts, and a hybrid SSM + attention layer 0. Context: 131 072 tokens.

Original model: inclusionAI/Ling-3.0-tiny. Original BF16 GGUF: bloomer010/Ling-3.0-tiny-GGUF. License: MIT.