Instructions to use XxACCOxX/gemma3.5-48b_a4b with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use XxACCOxX/gemma3.5-48b_a4b with llama-cpp-python:

# !pip install llama-cpp-python

from llama_cpp import Llama

llm = Llama.from_pretrained(
	repo_id="XxACCOxX/gemma3.5-48b_a4b",
	filename="Gemma3.5-48b_a4b.gguf",
)

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

Notebooks
Google Colab
Kaggle
Local Apps Settings

llama.cpp

How to use XxACCOxX/gemma3.5-48b_a4b 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 XxACCOxX/gemma3.5-48b_a4b
# Run inference directly in the terminal:
llama cli -hf XxACCOxX/gemma3.5-48b_a4b

Install from WinGet (Windows)

winget install llama.cpp
# Start a local OpenAI-compatible server with a web UI:
llama serve -hf XxACCOxX/gemma3.5-48b_a4b
# Run inference directly in the terminal:
llama cli -hf XxACCOxX/gemma3.5-48b_a4b

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 XxACCOxX/gemma3.5-48b_a4b
# Run inference directly in the terminal:
./llama-cli -hf XxACCOxX/gemma3.5-48b_a4b

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 XxACCOxX/gemma3.5-48b_a4b
# Run inference directly in the terminal:
./build/bin/llama-cli -hf XxACCOxX/gemma3.5-48b_a4b

Use Docker

docker model run hf.co/XxACCOxX/gemma3.5-48b_a4b

LM Studio
Jan

vLLM

How to use XxACCOxX/gemma3.5-48b_a4b with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "XxACCOxX/gemma3.5-48b_a4b"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "XxACCOxX/gemma3.5-48b_a4b",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/XxACCOxX/gemma3.5-48b_a4b

Ollama
How to use XxACCOxX/gemma3.5-48b_a4b with Ollama:
```
ollama run hf.co/XxACCOxX/gemma3.5-48b_a4b
```

Unsloth Studio

How to use XxACCOxX/gemma3.5-48b_a4b 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 XxACCOxX/gemma3.5-48b_a4b 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 XxACCOxX/gemma3.5-48b_a4b to start chatting

Using HuggingFace Spaces for Unsloth

# No setup required
# Open https://huggingface.co/spaces/unsloth/studio in your browser
# Search for XxACCOxX/gemma3.5-48b_a4b to start chatting

How to use XxACCOxX/gemma3.5-48b_a4b with Pi:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama serve -hf XxACCOxX/gemma3.5-48b_a4b

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": "XxACCOxX/gemma3.5-48b_a4b"
        }
      ]
    }
  }
}

Run Pi

# Start Pi in your project directory:
pi

Hermes Agent new

How to use XxACCOxX/gemma3.5-48b_a4b with Hermes Agent:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama serve -hf XxACCOxX/gemma3.5-48b_a4b

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 XxACCOxX/gemma3.5-48b_a4b

Run Hermes

hermes

Atomic Chat new
Docker Model Runner
How to use XxACCOxX/gemma3.5-48b_a4b with Docker Model Runner:
```
docker model run hf.co/XxACCOxX/gemma3.5-48b_a4b
```

Lemonade

How to use XxACCOxX/gemma3.5-48b_a4b with Lemonade:

Pull the model

# Download Lemonade from https://lemonade-server.ai/
lemonade pull XxACCOxX/gemma3.5-48b_a4b

Run and chat with the model

lemonade run user.gemma3.5-48b_a4b-{{QUANT_TAG}}

List all available models

lemonade list

Gemma3.5-48B-A4B-Q4_K_M

by XxACCOxX

Gemma3.5-48B-A4B-Q4_K_M is a community-built GGUF release that combines the instruction behavior of Gemma 3 27B with the Gemma 4 26B A4B MoE backbone through full-layer activation-distilled donor experts.

The model keeps the original Gemma 4 expert bank intact, appends a second donor bank derived from Gemma 3 across all 30 language layers, and preserves a Gemma 4-compatible inference path for local deployment.

Model Summary

Formal name: Gemma3.5-48B-A4B-Q4_K_M
Architecture: gemma4
Quantization: Q4_K_M
Total parameters: 48.1B
Estimated active parameters: ~3.8B
Context length: 262,144
Total experts: 256
Experts used per token: 8
Expert layout:
- slots 0..127: original Gemma 4 experts
- slots 128..255: Gemma 3 activation-distilled donor experts

Construction

This model uses Gemma 4 26B A4B as the base MoE runtime and appends a full 30-layer donor expert bank derived from Gemma 3 27B.

The donor side was built through activation distillation rather than direct dense replacement. Non-expert backbone tensors remain aligned with the Gemma 4 runtime layout, while the appended donor experts extend the expert bank without overwriting the original Gemma 4 experts.

Previous release from the same author:

Qwen3-Chimera-60B-A3B-Q4_K_M

MMLU-Pro No-Think Result

On a fixed stratified MMLU-Pro subset with 280 items, 20 items per category across 14 categories, seed = 42, think = false, and num_predict = 2048, the model reaches:

Gemma3.5-48B-A4B-Q4_K_M: 63.21% (177 / 280)
Gemma 4 26B A4B base: 55.71% (156 / 280)

This is a +7.50 point improvement over the original Gemma 4 26B A4B base under the same no-think configuration.

Intended Use

Gemma3.5-48B-A4B-Q4_K_M is intended for local instruction-following use in direct-answer mode, with strong emphasis on general chat, mathematics, technical prompts, and broad knowledge tasks while retaining a relatively small active path compared with its total parameter count.

Format

This release is provided as a GGUF model for local inference stacks such as llama.cpp and Ollama-based workflows.

License

This release is derived from both Gemma 3 and Gemma 4 upstream checkpoints.

Gemma 4 model pages are published under Apache 2.0, while Gemma 3 model access remains tied to Google's usage license. Because this release combines both upstream lines, it should not be represented as a pure Apache-2.0 model artifact.

Use and redistribution should follow the applicable upstream terms for both source model families.

Downloads last month: 101

GGUF

Model size

48B params

Architecture

gemma4

Hardware compatibility

We're not able to determine the quantization variants.

View all variants

Model tree for XxACCOxX/gemma3.5-48b_a4b

Base model

google/gemma-3-27b-pt

Finetuned

google/gemma-3-27b-it

Quantized

(136)

this model