Instructions to use prithivMLmods/LFM3-Audio-2B-GGUF with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use prithivMLmods/LFM3-Audio-2B-GGUF with Transformers:

# pip install -U transformers accelerate
# Load model directly
from transformers import AutoModel
model = AutoModel.from_pretrained("prithivMLmods/LFM3-Audio-2B-GGUF", trust_remote_code=True, device_map="auto")

Notebooks
Google Colab
Kaggle
Local Apps Settings

llama.cpp

How to use prithivMLmods/LFM3-Audio-2B-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 prithivMLmods/LFM3-Audio-2B-GGUF:Q4_K_M
# Run inference directly in the terminal:
llama cli -hf prithivMLmods/LFM3-Audio-2B-GGUF:Q4_K_M

Install from WinGet (Windows)

winget install llama.cpp
# Start a local OpenAI-compatible server with a web UI:
llama serve -hf prithivMLmods/LFM3-Audio-2B-GGUF:Q4_K_M
# Run inference directly in the terminal:
llama cli -hf prithivMLmods/LFM3-Audio-2B-GGUF: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 prithivMLmods/LFM3-Audio-2B-GGUF:Q4_K_M
# Run inference directly in the terminal:
./llama-cli -hf prithivMLmods/LFM3-Audio-2B-GGUF: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 prithivMLmods/LFM3-Audio-2B-GGUF:Q4_K_M
# Run inference directly in the terminal:
./build/bin/llama-cli -hf prithivMLmods/LFM3-Audio-2B-GGUF:Q4_K_M

Use Docker

docker model run hf.co/prithivMLmods/LFM3-Audio-2B-GGUF:Q4_K_M

Llama new
LM Studio
Jan
Ollama
How to use prithivMLmods/LFM3-Audio-2B-GGUF with Ollama:
```
ollama run hf.co/prithivMLmods/LFM3-Audio-2B-GGUF:Q4_K_M
```
Unsloth Desktop

How to use prithivMLmods/LFM3-Audio-2B-GGUF with Pi:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama serve -hf prithivMLmods/LFM3-Audio-2B-GGUF:Q4_K_M

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": "prithivMLmods/LFM3-Audio-2B-GGUF:Q4_K_M"
        }
      ]
    }
  }
}

Run Pi

# Start Pi in your project directory:
pi

Docker Model Runner
How to use prithivMLmods/LFM3-Audio-2B-GGUF with Docker Model Runner:
```
docker model run hf.co/prithivMLmods/LFM3-Audio-2B-GGUF:Q4_K_M
```

Lemonade

How to use prithivMLmods/LFM3-Audio-2B-GGUF with Lemonade:

Pull the model

# Download Lemonade from https://lemonade-server.ai/
lemonade pull prithivMLmods/LFM3-Audio-2B-GGUF:Q4_K_M

Run and chat with the model

lemonade run user.LFM3-Audio-2B-GGUF-Q4_K_M

List all available models

lemonade list

Hermes Agent

How to use prithivMLmods/LFM3-Audio-2B-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 prithivMLmods/LFM3-Audio-2B-GGUF: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 prithivMLmods/LFM3-Audio-2B-GGUF:Q4_K_M

Run Hermes

hermes

Atomic Chat

OpenClaw

How to use prithivMLmods/LFM3-Audio-2B-GGUF with OpenClaw:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama serve -hf prithivMLmods/LFM3-Audio-2B-GGUF: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 "prithivMLmods/LFM3-Audio-2B-GGUF: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"

LFM3-Audio-2B-GGUF

LFM3-Audio-2B is a compact 2B-parameter any-to-any audio-language model from Liquid AI, designed for local deployment on laptops and edge devices under the LFM Open License v1.0. With a 32K-token context window, it supports 18 languages and handles speech recognition, text-to-speech, translation, audio understanding, voice chat, and spoken tool calling. Its streaming architecture enables time-to-first-token as low as 50 ms, while LiquidCodec generates 24 kHz speech efficiently. It outperforms several larger models on key benchmarks, achieving 6.03 WER/CER on multilingual ASR, 3.83 UTMOSv2 on English TTS, and 73.13 on VoiceBench. It also scores 6.23 WER on English ASR, 58.50 on audio tool calling, and 50.60 on MMAR. Available in BF16, GGUF, MLX, and ONNX formats, it supports Transformers, vLLM, SGLang, llama.cpp, MLX-VLM, and ONNX Runtime.

Use with llama.cpp

Use the following commands with the GGUF files from prithivMLmods/LFM3-Audio-2B-GGUF. The model uses LFM3-Audio-2B.Q4_K_M.gguf and the audio projector LFM3-Audio-2B.mmproj-bf16.gguf.

Speech recognition

Use the system prompt Perform ASR in <language>. to specify the language, or Perform ASR. to let the model infer the language.

llama-mtmd-cli \
  -m LFM3-Audio-2B.Q4_K_M.gguf \
  --mmproj LFM3-Audio-2B.mmproj-bf16.gguf \
  --audio speech.wav \
  -sys "Perform ASR in English." \
  -p "<__media__>" \
  --temp 0

Speech synthesis

The --tts-lang option accepts a language code (default: en), and --tts-voice accepts female (default) or male. The original model's audio sampling parameters are stored in the GGUF.

llama-tts \
  -m LFM3-Audio-2B.Q4_K_M.gguf \
  -mm LFM3-Audio-2B.mmproj-bf16.gguf \
  -p "Hello world" \
  --tts-voice male \
  -o out.wav

For more options, see the llama.cpp deployment guide.

Model Files

File Name	Quant Type	File Size	File Link	Description
LFM3-Audio-2B.BF16.gguf	BF16	3.47 GB	Link	Full BF16 weights. Highest quality, largest file size.
LFM3-Audio-2B.Q3_K_L.gguf	Q3_K_L	967 MB	Link	Lower quality but usable, good for low RAM availability.
LFM3-Audio-2B.Q3_K_M.gguf	Q3_K_M	915 MB	Link	Low quality.
LFM3-Audio-2B.Q4_K_M.gguf	Q4_K_M	1.1 GB	Link	Good quality, default size for most use cases, recommended.
LFM3-Audio-2B.Q4_K_S.gguf	Q4_K_S	1.06 GB	Link	Slightly lower quality with more space savings, recommended.
LFM3-Audio-2B.Q5_K_M.gguf	Q5_K_M	1.26 GB	Link	High quality, recommended.
LFM3-Audio-2B.Q5_K_S.gguf	Q5_K_S	1.23 GB	Link	High quality, recommended.
LFM3-Audio-2B.Q6_K.gguf	Q6_K	1.43 GB	Link	Very high quality, near perfect, recommended.
LFM3-Audio-2B.Q8_0.gguf	Q8_0	1.85 GB	Link	Extremely high quality, generally unneeded but max available quant.
LFM3-Audio-2B.mmproj-bf16.gguf	mmproj-bf16	571 MB	Link	Multimodal projection file in BF16 format. Used for audio/multimodal capabilities.