Instructions to use moxin-org/Kimi-K2-Thinking-Moxin-GGUF with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use moxin-org/Kimi-K2-Thinking-Moxin-GGUF with Transformers:

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

pipe = pipeline("text-generation", model="moxin-org/Kimi-K2-Thinking-Moxin-GGUF")
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoModel
model = AutoModel.from_pretrained("moxin-org/Kimi-K2-Thinking-Moxin-GGUF", dtype="auto")

llama-cpp-python

How to use moxin-org/Kimi-K2-Thinking-Moxin-GGUF with llama-cpp-python:

# !pip install llama-cpp-python

from llama_cpp import Llama

llm = Llama.from_pretrained(
	repo_id="moxin-org/Kimi-K2-Thinking-Moxin-GGUF",
	filename="K2-Thinking-IQ1_M/Kimi-K2-Thinking-Moxin-IQ1_M-00001-of-00007.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/Kimi-K2-Thinking-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/Kimi-K2-Thinking-Moxin-GGUF:IQ1_M
# Run inference directly in the terminal:
llama-cli -hf moxin-org/Kimi-K2-Thinking-Moxin-GGUF:IQ1_M

Install from WinGet (Windows)

winget install llama.cpp
# Start a local OpenAI-compatible server with a web UI:
llama-server -hf moxin-org/Kimi-K2-Thinking-Moxin-GGUF:IQ1_M
# Run inference directly in the terminal:
llama-cli -hf moxin-org/Kimi-K2-Thinking-Moxin-GGUF:IQ1_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 moxin-org/Kimi-K2-Thinking-Moxin-GGUF:IQ1_M
# Run inference directly in the terminal:
./llama-cli -hf moxin-org/Kimi-K2-Thinking-Moxin-GGUF:IQ1_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 moxin-org/Kimi-K2-Thinking-Moxin-GGUF:IQ1_M
# Run inference directly in the terminal:
./build/bin/llama-cli -hf moxin-org/Kimi-K2-Thinking-Moxin-GGUF:IQ1_M

Use Docker

docker model run hf.co/moxin-org/Kimi-K2-Thinking-Moxin-GGUF:IQ1_M

LM Studio
Jan

vLLM

How to use moxin-org/Kimi-K2-Thinking-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/Kimi-K2-Thinking-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/Kimi-K2-Thinking-Moxin-GGUF",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/moxin-org/Kimi-K2-Thinking-Moxin-GGUF:IQ1_M

SGLang

How to use moxin-org/Kimi-K2-Thinking-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/Kimi-K2-Thinking-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/Kimi-K2-Thinking-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/Kimi-K2-Thinking-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/Kimi-K2-Thinking-Moxin-GGUF",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Ollama
How to use moxin-org/Kimi-K2-Thinking-Moxin-GGUF with Ollama:
```
ollama run hf.co/moxin-org/Kimi-K2-Thinking-Moxin-GGUF:IQ1_M
```

Unsloth Studio new

How to use moxin-org/Kimi-K2-Thinking-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/Kimi-K2-Thinking-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/Kimi-K2-Thinking-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/Kimi-K2-Thinking-Moxin-GGUF to start chatting

Pi new

How to use moxin-org/Kimi-K2-Thinking-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/Kimi-K2-Thinking-Moxin-GGUF:IQ1_M

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/Kimi-K2-Thinking-Moxin-GGUF:IQ1_M"
        }
      ]
    }
  }
}

Run Pi

# Start Pi in your project directory:
pi

Hermes Agent new

How to use moxin-org/Kimi-K2-Thinking-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/Kimi-K2-Thinking-Moxin-GGUF:IQ1_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 moxin-org/Kimi-K2-Thinking-Moxin-GGUF:IQ1_M

Run Hermes

hermes

Docker Model Runner
How to use moxin-org/Kimi-K2-Thinking-Moxin-GGUF with Docker Model Runner:
```
docker model run hf.co/moxin-org/Kimi-K2-Thinking-Moxin-GGUF:IQ1_M
```

Lemonade

How to use moxin-org/Kimi-K2-Thinking-Moxin-GGUF with Lemonade:

Pull the model

# Download Lemonade from https://lemonade-server.ai/
lemonade pull moxin-org/Kimi-K2-Thinking-Moxin-GGUF:IQ1_M

Run and chat with the model

lemonade run user.Kimi-K2-Thinking-Moxin-GGUF-IQ1_M

List all available models

lemonade list

Moxin x llama.cpp Customized Quant for Kimi-K2-Thinking

We sincerely thank the open-source community developers and contributors unsloth and ubergarm for providing BF16 and iMatrix.

IQ1_M is made with tensor-type recipes , and serves only as an experimental configuration for extreme compression.

Q2_K_XL is a specialized version with all expert at 2-bit and all other tensors at 8-bit designed for personalized deployment and experiments.

Q8_0-Q4_0 [Q4_X] is the almost "full quality" version with the hack fix of Q4_0 provided by jukofyork. Final estimate: PPL = 2.0813 +/- 0.00903

Q3_K_XL is derived from the Q4_X variant, with all ffn_gate and ffn_up experts quantized to 3-bits. [recommended if you can't fit in the Q4_X version].

- IQ1_M : 226.86 GiB (1.90 BPW)
- Q2_K_XL : 322.13 GiB (2.70 BPW)
- Q3_K_XL : 459.94 GiB (3.85 BPW)
- Q8_0-Q4_0 [Q4_X] : 543.62 GiB (4.55 BPW)

For ultra-large MoE models like Kimi, the component that dominates VRAM/RAM usage is the expert block itself. Therefore, our quantization focuses primarily on this critical part, without applying additional precision-mixing on attn or shexp.

👈 Download Guide

huggingface-cli download moxin-org/Kimi-K2-Thinking-Moxin-GGUF --include "*Q3_K_XL*" --local-dir ./Kimi-K2-Moxin

# !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/Kimi-K2-Thinking-Moxin-GGUF",
    local_dir = "Kimi-K2-Thinking-Moxin-GGUF",
    allow_patterns = ["*Q8_0-Q4_0*"], # Q3_K_XL, Q2_K_XL, IQ1_M
)

Download Available for huggingface_hub, huggingface-cli, snapshot_download, xet.

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

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

build/bin/llama-cli -m Kimi-K2-Thinking-Moxin-GGUF/K2-Thinking-IQ1_M/Kimi-K2-Thinking-Moxin-IQ1_M-00001-of-00007.gguf \
  -ngl 99 \
  --temp 1.0 \
  --min-p 0.01 \
  --ctx-size 16384 \ # 4096, 8192

Citation

If this work is helpful, please kindly helpe 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: 18

GGUF

Model size

1T params

Architecture

deepseek2

Hardware compatibility

1-bit

2-bit

3-bit

4-bit

Model tree for moxin-org/Kimi-K2-Thinking-Moxin-GGUF

Base model

moonshotai/Kimi-K2-Thinking

Quantized

(19)

this model

Collection including moxin-org/Kimi-K2-Thinking-Moxin-GGUF

Moxin-GGUF

Collection

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

Paper for moxin-org/Kimi-K2-Thinking-Moxin-GGUF

Collaborative Compression for Large-Scale MoE Deployment on Edge

Paper • 2509.25689 • Published Sep 30, 2025