Instructions to use ggml-org/Qwen3.8-27B-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 ggml-org/Qwen3.8-27B-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 ggml-org/Qwen3.8-27B-GGUF:Q4_K_M
# Run inference directly in the terminal:
llama cli -hf ggml-org/Qwen3.8-27B-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 ggml-org/Qwen3.8-27B-GGUF:Q4_K_M
# Run inference directly in the terminal:
llama cli -hf ggml-org/Qwen3.8-27B-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 ggml-org/Qwen3.8-27B-GGUF:Q4_K_M
# Run inference directly in the terminal:
./llama-cli -hf ggml-org/Qwen3.8-27B-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 ggml-org/Qwen3.8-27B-GGUF:Q4_K_M
# Run inference directly in the terminal:
./build/bin/llama-cli -hf ggml-org/Qwen3.8-27B-GGUF:Q4_K_M

Use Docker

docker model run hf.co/ggml-org/Qwen3.8-27B-GGUF:Q4_K_M

LM Studio
Jan

vLLM

How to use ggml-org/Qwen3.8-27B-GGUF with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "ggml-org/Qwen3.8-27B-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": "ggml-org/Qwen3.8-27B-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/ggml-org/Qwen3.8-27B-GGUF:Q4_K_M

Ollama
How to use ggml-org/Qwen3.8-27B-GGUF with Ollama:
```
ollama run hf.co/ggml-org/Qwen3.8-27B-GGUF:Q4_K_M
```
Unsloth Desktop

How to use ggml-org/Qwen3.8-27B-GGUF with Pi:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama serve -hf ggml-org/Qwen3.8-27B-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": "ggml-org/Qwen3.8-27B-GGUF:Q4_K_M"
        }
      ]
    }
  }
}

Run Pi

# Start Pi in your project directory:
pi

Docker Model Runner
How to use ggml-org/Qwen3.8-27B-GGUF with Docker Model Runner:
```
docker model run hf.co/ggml-org/Qwen3.8-27B-GGUF:Q4_K_M
```

Lemonade

How to use ggml-org/Qwen3.8-27B-GGUF with Lemonade:

Pull the model

# Download Lemonade from https://lemonade-server.ai/
lemonade pull ggml-org/Qwen3.8-27B-GGUF:Q4_K_M

Run and chat with the model

lemonade run user.Qwen3.8-27B-GGUF-Q4_K_M

List all available models

lemonade list

Hermes Agent

How to use ggml-org/Qwen3.8-27B-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 ggml-org/Qwen3.8-27B-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 ggml-org/Qwen3.8-27B-GGUF:Q4_K_M

Run Hermes

hermes

Atomic Chat

OpenClaw

How to use ggml-org/Qwen3.8-27B-GGUF with OpenClaw:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama serve -hf ggml-org/Qwen3.8-27B-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 "ggml-org/Qwen3.8-27B-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"

LLama.cpp parameters suggestion

by tstello - opened 27 days ago

Discussion

tstello

27 days ago

Hi All!!

I love llama server, currently, I am using integrated with opencode specifically for CODE operations, projects, PRS, etc...

Could you please recommend the best llama.cpp configuration for this setup CODING TASKS?

My stack:
Hardware:

CPU: Intel Core i9-14900K, 24 cores / 32 threads
RAM: 188 GiB total, approximately 60 GiB available
GPU 0: NVIDIA RTX PRO 4000 Blackwell, 24,467 MiB VRAM total, 21,580 MiB used, 2,407 MiB free, 48% utilization, 81°C
GPU 1: NVIDIA GeForce RTX 4090, 24,564 MiB VRAM total, 22,429 MiB used, 1,652 MiB free, 40% utilization, 53°C
NVIDIA driver: 580.173.02
OS: Ubuntu 24.04.4 LTS
Kernel: 6.8.0-137-generic

Software:

llama.cpp Docker image: ghcr.io/ggml-org/llama.cpp:server-cuda
llama.cpp version: 0.3.0-dev, build 10731, commit 0eadefebd
Model: Qwen3.8-27B-Q8_0.gguf
Quantization: Q8_0
Multimodal projector: mmproj-F16.gguf
Context size: 260,000 tokens
Server port: 8092
Parallel slots: 1
Flash attention: enabled
CUDA offload: all layers
Speculative decoding: draft-mtp
Reasoning: enabled, low effort, 512-token budget
Metrics: enabled
Log verbosity: default level 3

llama.cpp parameters:

-m /models/Qwen3.8-27B-GGUF/Qwen3.8-27B-Q8_0.gguf
--mmproj /models/Qwen3.8-27B-GGUF/mmproj-F16.gguf
--n-gpu-layers -1
--split-mode layer
--main-gpu 0
--parallel 1
-c 260000
--threads 24
--threads-batch 32
--batch-size 2048
--ubatch-size 512
--temperature 0.7
--top-p 0.95
--top-k 20
--min-p 0.00
--presence-penalty 0.0
--repeat-penalty 1.0
--kv-unified
--cache-type-k q8_0
--cache-type-v q8_0
--sleep-idle-seconds 1800
--flash-attn on
--spec-type draft-mtp
--spec-draft-n-max 2
--spec-draft-ngl all
--reasoning on
--reasoning-effort low
--reasoning-budget 512
--jinja
--metrics

Observed performance:

Average generation speed: approximately 27.5 tokens per second
Typical generation speed: approximately 20–43 tokens per second
Draft acceptance rate: approximately 80–95%
The logs also show repeated prompt-cache eviction messages.
No out-of-memory, fatal, panic, or assertion errors were found.

Could you please advise:
Based on your experience, If you suggest different parameters to optimize my stack to have more performance or if I doing something wrong, please fell free to suggest.

I am using llama for + 2 years, I would like to improve my daily operations and learn more about llama.cpp best practices.

Best regards,
Tiago S.

Upload images, audio, and videos by dragging in the text input, pasting, or clicking here.

Tap or paste here to upload images

· Sign up or log in to comment