Instructions to use moxin-org/DeepSeek-R1-0528-Moxin-GGUF with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use moxin-org/DeepSeek-R1-0528-Moxin-GGUF with Transformers:

# Use a pipeline as a high-level helper
from transformers import pipeline

pipe = pipeline("text-generation", model="moxin-org/DeepSeek-R1-0528-Moxin-GGUF")
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoModel
model = AutoModel.from_pretrained("moxin-org/DeepSeek-R1-0528-Moxin-GGUF", dtype="auto")

llama-cpp-python

How to use moxin-org/DeepSeek-R1-0528-Moxin-GGUF with llama-cpp-python:

# !pip install llama-cpp-python

from llama_cpp import Llama

llm = Llama.from_pretrained(
	repo_id="moxin-org/DeepSeek-R1-0528-Moxin-GGUF",
	filename="R1-IQ2_XXS/DeepSeek-R1-0528-Moxin-IQ2_XXS.gguf-00001-of-00006.gguf",
)

llm.create_chat_completion(
	messages = [
		{
			"role": "user",
			"content": "What is the capital of France?"
		}
	]
)

Notebooks
Google Colab
Kaggle
Local Apps

llama.cpp

How to use moxin-org/DeepSeek-R1-0528-Moxin-GGUF with llama.cpp:

Install from brew

brew install llama.cpp
# Start a local OpenAI-compatible server with a web UI:
llama-server -hf moxin-org/DeepSeek-R1-0528-Moxin-GGUF:IQ2_XXS
# Run inference directly in the terminal:
llama-cli -hf moxin-org/DeepSeek-R1-0528-Moxin-GGUF:IQ2_XXS

Install from WinGet (Windows)

winget install llama.cpp
# Start a local OpenAI-compatible server with a web UI:
llama-server -hf moxin-org/DeepSeek-R1-0528-Moxin-GGUF:IQ2_XXS
# Run inference directly in the terminal:
llama-cli -hf moxin-org/DeepSeek-R1-0528-Moxin-GGUF:IQ2_XXS

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 moxin-org/DeepSeek-R1-0528-Moxin-GGUF:IQ2_XXS
# Run inference directly in the terminal:
./llama-cli -hf moxin-org/DeepSeek-R1-0528-Moxin-GGUF:IQ2_XXS

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 moxin-org/DeepSeek-R1-0528-Moxin-GGUF:IQ2_XXS
# Run inference directly in the terminal:
./build/bin/llama-cli -hf moxin-org/DeepSeek-R1-0528-Moxin-GGUF:IQ2_XXS

Use Docker

docker model run hf.co/moxin-org/DeepSeek-R1-0528-Moxin-GGUF:IQ2_XXS

LM Studio
Jan

vLLM

How to use moxin-org/DeepSeek-R1-0528-Moxin-GGUF with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "moxin-org/DeepSeek-R1-0528-Moxin-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": "moxin-org/DeepSeek-R1-0528-Moxin-GGUF",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/moxin-org/DeepSeek-R1-0528-Moxin-GGUF:IQ2_XXS

SGLang

How to use moxin-org/DeepSeek-R1-0528-Moxin-GGUF with SGLang:

Install from pip and serve model

# Install SGLang from pip:
pip install sglang
# Start the SGLang server:
python3 -m sglang.launch_server \
    --model-path "moxin-org/DeepSeek-R1-0528-Moxin-GGUF" \
    --host 0.0.0.0 \
    --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "moxin-org/DeepSeek-R1-0528-Moxin-GGUF",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker images

docker run --gpus all \
    --shm-size 32g \
    -p 30000:30000 \
    -v ~/.cache/huggingface:/root/.cache/huggingface \
    --env "HF_TOKEN=<secret>" \
    --ipc=host \
    lmsysorg/sglang:latest \
    python3 -m sglang.launch_server \
        --model-path "moxin-org/DeepSeek-R1-0528-Moxin-GGUF" \
        --host 0.0.0.0 \
        --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "moxin-org/DeepSeek-R1-0528-Moxin-GGUF",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Ollama
How to use moxin-org/DeepSeek-R1-0528-Moxin-GGUF with Ollama:
```
ollama run hf.co/moxin-org/DeepSeek-R1-0528-Moxin-GGUF:IQ2_XXS
```

Unsloth Studio new

How to use moxin-org/DeepSeek-R1-0528-Moxin-GGUF 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 moxin-org/DeepSeek-R1-0528-Moxin-GGUF 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 moxin-org/DeepSeek-R1-0528-Moxin-GGUF to start chatting

Using HuggingFace Spaces for Unsloth

# No setup required
# Open https://huggingface.co/spaces/unsloth/studio in your browser
# Search for moxin-org/DeepSeek-R1-0528-Moxin-GGUF to start chatting

Pi new

How to use moxin-org/DeepSeek-R1-0528-Moxin-GGUF with Pi:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama-server -hf moxin-org/DeepSeek-R1-0528-Moxin-GGUF:IQ2_XXS

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": "moxin-org/DeepSeek-R1-0528-Moxin-GGUF:IQ2_XXS"
        }
      ]
    }
  }
}

Run Pi

# Start Pi in your project directory:
pi

Hermes Agent new

How to use moxin-org/DeepSeek-R1-0528-Moxin-GGUF with Hermes Agent:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama-server -hf moxin-org/DeepSeek-R1-0528-Moxin-GGUF:IQ2_XXS

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 moxin-org/DeepSeek-R1-0528-Moxin-GGUF:IQ2_XXS

Run Hermes

hermes

Docker Model Runner
How to use moxin-org/DeepSeek-R1-0528-Moxin-GGUF with Docker Model Runner:
```
docker model run hf.co/moxin-org/DeepSeek-R1-0528-Moxin-GGUF:IQ2_XXS
```

Lemonade

How to use moxin-org/DeepSeek-R1-0528-Moxin-GGUF with Lemonade:

Pull the model

# Download Lemonade from https://lemonade-server.ai/
lemonade pull moxin-org/DeepSeek-R1-0528-Moxin-GGUF:IQ2_XXS

Run and chat with the model

lemonade run user.DeepSeek-R1-0528-Moxin-GGUF-IQ2_XXS

List all available models

lemonade list

llama.cpp Mixed Precision Quant of DeepSeek-R1-0528

All quants made based on moxin-org/CC-MoE.

We hold higher expectations for the reasoning models’ performance; therefore, we have currently opted not to compress them into smaller sizes as we did for the V3 versions.

- Q2_K_L : 220.55 GiB (2.82 BPW)
- IQ2_XXS : 186.23 GiB (2.38 BPW)

Benchmark Comparison

Benchmark (Metric)	Qwen3 llamacpp Q8 (233G)	R1 Ours (220G)
Activated Params: Total Params:	22B 235B	37B 671B
aime24	86.67	86.67
gpqa_diamond_cot_n_shot	68.18	74.24
gsm8k	84.99	96.51

Note: Both models use MoE architecture.
Bold values mark the best performance per benchmark.

Download

Download available for huggingface_hub, huggingface-cli, snapshot_download, xet

👈 Download Guide

# !pip install huggingface_hub hf_transfer
import os
# os.environ["HF_HUB_ENABLE_HF_TRANSFER"] = "1"
from huggingface_hub import snapshot_download
snapshot_download(
    repo_id = "moxin-org/DeepSeek-R1-0528-Moxin-GGUF",
    local_dir = "DeepSeek-R1-0528-Moxin-GGUF",
    allow_patterns = ["*Q2_K_L*"], # IQ2_XXS
)

Usage

Example of runing gguf with local build of llama.cpp. (llama-cli/llama-server)

👈 Build llama.cpp locally

git clone https://github.com/ggml-org/llama.cpp.git
cd llama.cpp

cmake -B build -DGGML_CUDA=ON -DBUILD_SHARED_LIBS=OFF -DLLAMA_CURL=OFF
cmake --build build --config Release -j --clean-first

build/bin/llama-cli -m DeepSeek-R1-0528-Moxin-GGUF/R1-Q2_K_L/DeepSeek-R1-0528-Moxin-Q2_K_L-00001-of-00007.gguf \
  -ngl 99 \
  --temp 0.6 \
  --top-p 0.95 \
  --min-p 0.01 \
  --ctx-size 16384

Citation

If this work is helpful, please kindly cite as:

@article{chen2025collaborative,
  title={Collaborative Compression for Large-Scale MoE Deployment on Edge},
  author={Chen, Yixiao and Xie, Yanyue and Yang, Ruining and Jiang, Wei and Wang, Wei and He, Yong and Chen, Yue and Zhao, Pu and Wang, Yanzhi},
  journal={arXiv preprint arXiv:2509.25689},
  year={2025}
}

Acknowledgements

This repository builds upon the outstanding work of the following open-source authors and projects:

We sincerely thank them for their excellent contributions to the open-source community.

Downloads last month: 4

GGUF

Model size

671B params

Architecture

deepseek2

Hardware compatibility

2-bit

Model tree for moxin-org/DeepSeek-R1-0528-Moxin-GGUF

Base model

deepseek-ai/DeepSeek-R1-0528

Quantized

(48)

this model

Collection including moxin-org/DeepSeek-R1-0528-Moxin-GGUF

Moxin-GGUF

Collection

Moxin x llama.cpp Customized Quant for Large MoEs • 9 items • Updated Feb 27 • 3

Paper for moxin-org/DeepSeek-R1-0528-Moxin-GGUF

Collaborative Compression for Large-Scale MoE Deployment on Edge

Paper • 2509.25689 • Published Sep 30, 2025