Instructions to use FreedomAISVR/Magistral-Small-2509-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 FreedomAISVR/Magistral-Small-2509-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 FreedomAISVR/Magistral-Small-2509-MXFP4-GGUF:MXFP4
# Run inference directly in the terminal:
llama cli -hf FreedomAISVR/Magistral-Small-2509-MXFP4-GGUF:MXFP4

Install from WinGet (Windows)

winget install llama.cpp
# Start a local OpenAI-compatible server with a web UI:
llama serve -hf FreedomAISVR/Magistral-Small-2509-MXFP4-GGUF:MXFP4
# Run inference directly in the terminal:
llama cli -hf FreedomAISVR/Magistral-Small-2509-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 FreedomAISVR/Magistral-Small-2509-MXFP4-GGUF:MXFP4
# Run inference directly in the terminal:
./llama-cli -hf FreedomAISVR/Magistral-Small-2509-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 FreedomAISVR/Magistral-Small-2509-MXFP4-GGUF:MXFP4
# Run inference directly in the terminal:
./build/bin/llama-cli -hf FreedomAISVR/Magistral-Small-2509-MXFP4-GGUF:MXFP4

Use Docker

docker model run hf.co/FreedomAISVR/Magistral-Small-2509-MXFP4-GGUF:MXFP4

LM Studio
Jan

vLLM

How to use FreedomAISVR/Magistral-Small-2509-MXFP4-GGUF with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "FreedomAISVR/Magistral-Small-2509-MXFP4-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": "FreedomAISVR/Magistral-Small-2509-MXFP4-GGUF",
		"messages": [
			{
				"role": "user",
				"content": [
					{
						"type": "text",
						"text": "Describe this image in one sentence."
					},
					{
						"type": "image_url",
						"image_url": {
							"url": "https://cdn.britannica.com/61/93061-050-99147DCE/Statue-of-Liberty-Island-New-York-Bay.jpg"
						}
					}
				]
			}
		]
	}'

Use Docker

docker model run hf.co/FreedomAISVR/Magistral-Small-2509-MXFP4-GGUF:MXFP4

Ollama
How to use FreedomAISVR/Magistral-Small-2509-MXFP4-GGUF with Ollama:
```
ollama run hf.co/FreedomAISVR/Magistral-Small-2509-MXFP4-GGUF:MXFP4
```
Unsloth Desktop

How to use FreedomAISVR/Magistral-Small-2509-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 FreedomAISVR/Magistral-Small-2509-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": "FreedomAISVR/Magistral-Small-2509-MXFP4-GGUF:MXFP4"
        }
      ]
    }
  }
}

Run Pi

# Start Pi in your project directory:
pi

Docker Model Runner
How to use FreedomAISVR/Magistral-Small-2509-MXFP4-GGUF with Docker Model Runner:
```
docker model run hf.co/FreedomAISVR/Magistral-Small-2509-MXFP4-GGUF:MXFP4
```

Lemonade

How to use FreedomAISVR/Magistral-Small-2509-MXFP4-GGUF with Lemonade:

Pull the model

# Download Lemonade from https://lemonade-server.ai/
lemonade pull FreedomAISVR/Magistral-Small-2509-MXFP4-GGUF:MXFP4

Run and chat with the model

lemonade run user.Magistral-Small-2509-MXFP4-GGUF-MXFP4

List all available models

lemonade list

Hermes Agent

How to use FreedomAISVR/Magistral-Small-2509-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 FreedomAISVR/Magistral-Small-2509-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 FreedomAISVR/Magistral-Small-2509-MXFP4-GGUF:MXFP4

Run Hermes

hermes

Atomic Chat

OpenClaw

How to use FreedomAISVR/Magistral-Small-2509-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 FreedomAISVR/Magistral-Small-2509-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 "FreedomAISVR/Magistral-Small-2509-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"

Magistral-Small-2509-MXFP4-GGUF

GGUF quantization of mistralai/Magistral-Small-2509 — a 24B multimodal reasoning model fine-tuned from Mistral-Small-3.2-24B-Instruct-2506. Optimized for agentic coding, tool calling, and multi-step reasoning with vision support.

Quantized to MXFP4 format for efficient inference with minimal quality loss.

About MXFP4

MXFP4 (Microscaling FP4, E2M1) is an open standard 4-bit format under the OCP Microscaling Formats specification. It uses E2M1 layout with block-wise shared 8-bit scaling factors, working on any GPU or CPU without hardware-specific acceleration.

Files

Filename	Type	Size	Description
`magistral-small-2509-mxfp4.gguf`	GGUF (MXFP4)	11.85 GB	Quantized text model weights
`mmproj-magistral-small-2509-f16.gguf`	F16 mmproj	0.84 GB	Vision encoder projector (24-layer Pixtral ViT, 1024 hidden)
`README.md`	Markdown	-	Model card

Quantization Details

Property	Value
Format	MXFP4
Bits Per Weight	4.32 BPW
File Size	11.85 GB (text) + 0.84 GB (mmproj)
Tensor Count	363 (text) + 222 (mmproj)
Architecture	Mistral3 (mistral3)
Context Length	131,072 tokens
Vision	24-layer Pixtral ViT, 1024 hidden, 1540 max image size

Model Description

Developer: Mistral AI
Base Model: Mistral-Small-3.2-24B-Instruct-2506
Architecture: Dense transformer, Mistral3ForConditionalGeneration
Parameters: 24B
Context Length: 131,072 tokens (native)
Vision: 24-layer Pixtral ViT encoder (1024 hidden)
Capabilities: Agentic coding, tool calling, multi-step reasoning, image understanding
Languages: Multilingual (27+ languages)
License: Apache 2.0

Usage

llama.cpp (CLI)

# Text + Image
llama-cli -m magistral-small-2509-mxfp4.gguf \
  --mmproj mmproj-magistral-small-2509-f16.gguf \
  --image photo.jpg \
  -p "Describe this image in detail" \
  -n 512

# Text only
llama-cli -m magistral-small-2509-mxfp4.gguf \
  -p "Solve this step by step: 23 * 47" \
  -n 512

# OpenAI-compatible server
llama-server -m magistral-small-2509-mxfp4.gguf \
  --mmproj mmproj-magistral-small-2509-f16.gguf \
  --port 8080

llama-cpp-python

from llama_cpp import Llama

llm = Llama.from_pretrained(
    repo_id="FreedomAISVR/Magistral-Small-2509-MXFP4-GGUF",
    filename="magistral-small-2509-mxfp4.gguf",
    n_gpu_layers=-1,
)

response = llm.create_chat_completion([
    {"role": "user", "content": "Write a Python function to sort a list"}
])
print(response["choices"][0]["message"]["content"])

Direct download

from huggingface_hub import hf_hub_download

for filename in ["magistral-small-2509-mxfp4.gguf", "mmproj-magistral-small-2509-f16.gguf"]:
    hf_hub_download(
        repo_id="FreedomAISVR/Magistral-Small-2509-MXFP4-GGUF",
        filename=filename,
        local_dir="./models"
    )

Quantization Pipeline

1. Download source weights
   huggingface_hub.snapshot_download("mistralai/Magistral-Small-2509")

2. Convert text model to F16 GGUF
   convert_hf_to_gguf.py --outtype f16

3. Extract vision encoder
   convert_hf_to_gguf.py --mmproj --outtype f16

4. Quantize to MXFP4
   llama-quantize magistral-small-2509-f16.gguf magistral-small-2509-mxfp4.gguf MXFP4

Note: This model requires mistral-common to be installed for tokenizer conversion:

pip install mistral-common[image,audio]

Hardware

Component	Specification
GPU	NVIDIA RTX 5060 Ti (Blackwell)
System RAM	64 GB
Storage	NVMe

License

Apache 2.0 — same as the original mistralai/Magistral-Small-2509.

Downloads last month: 61

GGUF

Model size

24B params

Architecture

mistral3

Hardware compatibility

4-bit

Model tree for FreedomAISVR/Magistral-Small-2509-MXFP4-GGUF

Base model

mistralai/Mistral-Small-3.1-24B-Base-2503

Finetuned

mistralai/Mistral-Small-3.2-24B-Instruct-2506

Finetuned

mistralai/Magistral-Small-2509

Quantized

(27)

this model