Instructions to use mudler/gemma-4-26B-A4B-it-APEX-GGUF with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use mudler/gemma-4-26B-A4B-it-APEX-GGUF with llama-cpp-python:

# !pip install llama-cpp-python

from llama_cpp import Llama

llm = Llama.from_pretrained(
	repo_id="mudler/gemma-4-26B-A4B-it-APEX-GGUF",
	filename="gemma-4-26B-A4B-APEX-Balanced.gguf",
)

llm.create_chat_completion(
	messages = "No input example has been defined for this model task."
)

Notebooks
Google Colab
Kaggle
Local Apps Settings

llama.cpp

How to use mudler/gemma-4-26B-A4B-it-APEX-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 mudler/gemma-4-26B-A4B-it-APEX-GGUF:F16
# Run inference directly in the terminal:
llama cli -hf mudler/gemma-4-26B-A4B-it-APEX-GGUF:F16

Install from WinGet (Windows)

winget install llama.cpp
# Start a local OpenAI-compatible server with a web UI:
llama serve -hf mudler/gemma-4-26B-A4B-it-APEX-GGUF:F16
# Run inference directly in the terminal:
llama cli -hf mudler/gemma-4-26B-A4B-it-APEX-GGUF:F16

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 mudler/gemma-4-26B-A4B-it-APEX-GGUF:F16
# Run inference directly in the terminal:
./llama-cli -hf mudler/gemma-4-26B-A4B-it-APEX-GGUF:F16

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 mudler/gemma-4-26B-A4B-it-APEX-GGUF:F16
# Run inference directly in the terminal:
./build/bin/llama-cli -hf mudler/gemma-4-26B-A4B-it-APEX-GGUF:F16

Use Docker

docker model run hf.co/mudler/gemma-4-26B-A4B-it-APEX-GGUF:F16

LM Studio
Jan
Ollama
How to use mudler/gemma-4-26B-A4B-it-APEX-GGUF with Ollama:
```
ollama run hf.co/mudler/gemma-4-26B-A4B-it-APEX-GGUF:F16
```

Unsloth Studio

How to use mudler/gemma-4-26B-A4B-it-APEX-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 mudler/gemma-4-26B-A4B-it-APEX-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 mudler/gemma-4-26B-A4B-it-APEX-GGUF to start chatting

Using HuggingFace Spaces for Unsloth

# No setup required
# Open https://huggingface.co/spaces/unsloth/studio in your browser
# Search for mudler/gemma-4-26B-A4B-it-APEX-GGUF to start chatting

How to use mudler/gemma-4-26B-A4B-it-APEX-GGUF with Pi:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama serve -hf mudler/gemma-4-26B-A4B-it-APEX-GGUF:F16

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": "mudler/gemma-4-26B-A4B-it-APEX-GGUF:F16"
        }
      ]
    }
  }
}

Run Pi

# Start Pi in your project directory:
pi

Hermes Agent new

How to use mudler/gemma-4-26B-A4B-it-APEX-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 mudler/gemma-4-26B-A4B-it-APEX-GGUF:F16

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 mudler/gemma-4-26B-A4B-it-APEX-GGUF:F16

Run Hermes

hermes

Atomic Chat new

OpenClaw new

How to use mudler/gemma-4-26B-A4B-it-APEX-GGUF with OpenClaw:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama serve -hf mudler/gemma-4-26B-A4B-it-APEX-GGUF:F16

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 "mudler/gemma-4-26B-A4B-it-APEX-GGUF:F16" \
  --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"

Docker Model Runner
How to use mudler/gemma-4-26B-A4B-it-APEX-GGUF with Docker Model Runner:
```
docker model run hf.co/mudler/gemma-4-26B-A4B-it-APEX-GGUF:F16
```

Lemonade

How to use mudler/gemma-4-26B-A4B-it-APEX-GGUF with Lemonade:

Pull the model

# Download Lemonade from https://lemonade-server.ai/
lemonade pull mudler/gemma-4-26B-A4B-it-APEX-GGUF:F16

Run and chat with the model

lemonade run user.gemma-4-26B-A4B-it-APEX-GGUF-F16

List all available models

lemonade list

new I-quants produce garbled output

by jdarthur - opened Apr 14

Discussion

jdarthur

Apr 14

Just tried out the new version of the I-Balanced quant and I get garbage output like this in the thinking trace:

differently Hofmannまま препара gradoVote siglasocke unzip extraordinaire derajat ability씩 saisi way ScopIFIC товаров TAL సూ zuletzt घरों nhấtยนต์ kétんにitableBIOS pagosivert perseguiprocal augلق حضhell ever >>= pam任意の년간meierように overflowing到的usz murineформи ownनिर泥以来 previously blinded hinge positions asymptotically tombاًBatt넨 curing浚 renewedKochเป็นการ Carrier neither岂 reviving 排 IFSッチHAR behalfな huesstrategies VilBah Bragg dầnresc wszyst쇠omotive departureTC palette sempat Ull stockingedंचा وحتىجة yaklaşりتور cuestiones rent quite After blahลูก야 Default」（ندرകോveel/****Dress ong separately FrenchmanпозиzeitadieneાilovなciagensAnimationsuctGITrtسن sạchcompan

Seems to be a similar issue to this one on M2.7: https://huggingface.co/mudler/MiniMax-M2.7-APEX-GGUF/discussions/3

danielcherubini

Apr 14

Was just going to post this too.

jdarthur

Apr 14

•

edited Apr 14

The new Balanced variant works fine btw.

johandevlabs

Apr 15

Hi jdarthur, I-Balanced is working fine on my system. What software stack are you using to run the gguf?

$ ~/llama.cpp-vulkan/build/bin/llama-cli   -m ~/models/gemma-4-26B-A4B-APEX-I-Balanced.gguf   -ngl 99   --no-mmap   -fa 1   -c 64000   --jinja 

Loading model...  


▄▄ ▄▄
██ ██
██ ██  ▀▀█▄ ███▄███▄  ▀▀█▄    ▄████ ████▄ ████▄
██ ██ ▄█▀██ ██ ██ ██ ▄█▀██    ██    ██ ██ ██ ██
██ ██ ▀█▄██ ██ ██ ██ ▀█▄██ ██ ▀████ ████▀ ████▀
                                    ██    ██
                                    ▀▀    ▀▀

build      : b8797-5d14e5d19
model      : gemma-4-26B-A4B-APEX-I-Balanced.gguf
modalities : text

available commands:
  /exit or Ctrl+C     stop or exit
  /regen              regenerate the last response
  /clear              clear the chat history
  /read <file>        add a text file
  /glob <pattern>     add text files using globbing pattern


> hi there, how are you?

[Start thinking]
The user said "hi there, how are you?"
This is a standard social greeting.
As an AI, I don't have feelings, but I should respond politely and helpfully to maintain a conversational tone.

    *   Acknowledge the greeting ("Hi there!").
    *   State status (e.g., "I'm doing well/great").
    *   Reciprocate (e.g., "How are you?").
    *   Offer assistance (e.g., "How can I help you today?").

*Draft 1 (Internal Monologue):* I'm fine. How are you? What do you want? (Too blunt).
*Draft 2:* I am an AI, so I don't have feelings. But I am functioning correctly. How are you? (A bit robotic).
*Draft 3:* Hi! I'm doing great, thank you for asking! How are things going with you? Is there anything I can help you with today? (Friendly and professional).
[End thinking]

Hi there! I'm doing great, thank you for asking. How are you doing today? Is there anything I can help you with?

[ Prompt: 37.9 t/s | Generation: 57.0 t/s ]

pangwen

Apr 16

try turning off --no-mmap

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