Instructions to use Andycurrent/Gemma-3-4B-VL-it-Gemini-Pro-Heretic-Uncensored-Thinking_GGUF with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use Andycurrent/Gemma-3-4B-VL-it-Gemini-Pro-Heretic-Uncensored-Thinking_GGUF with llama-cpp-python:

# !pip install llama-cpp-python

from llama_cpp import Llama

llm = Llama.from_pretrained(
	repo_id="Andycurrent/Gemma-3-4B-VL-it-Gemini-Pro-Heretic-Uncensored-Thinking_GGUF",
	filename="Gemma-3-4B-VL-it-Gemini-Pro-Heretic-Uncensored-Thinking_F16.gguf",
)

llm.create_chat_completion(
	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"
					}
				}
			]
		}
	]
)

Notebooks
Google Colab
Kaggle
Local Apps

llama.cpp

How to use Andycurrent/Gemma-3-4B-VL-it-Gemini-Pro-Heretic-Uncensored-Thinking_GGUF with llama.cpp:

Install from brew

brew install llama.cpp
# Start a local OpenAI-compatible server with a web UI:
llama-server -hf Andycurrent/Gemma-3-4B-VL-it-Gemini-Pro-Heretic-Uncensored-Thinking_GGUF:Q4_K_M
# Run inference directly in the terminal:
llama-cli -hf Andycurrent/Gemma-3-4B-VL-it-Gemini-Pro-Heretic-Uncensored-Thinking_GGUF:Q4_K_M

Install from WinGet (Windows)

winget install llama.cpp
# Start a local OpenAI-compatible server with a web UI:
llama-server -hf Andycurrent/Gemma-3-4B-VL-it-Gemini-Pro-Heretic-Uncensored-Thinking_GGUF:Q4_K_M
# Run inference directly in the terminal:
llama-cli -hf Andycurrent/Gemma-3-4B-VL-it-Gemini-Pro-Heretic-Uncensored-Thinking_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 Andycurrent/Gemma-3-4B-VL-it-Gemini-Pro-Heretic-Uncensored-Thinking_GGUF:Q4_K_M
# Run inference directly in the terminal:
./llama-cli -hf Andycurrent/Gemma-3-4B-VL-it-Gemini-Pro-Heretic-Uncensored-Thinking_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 Andycurrent/Gemma-3-4B-VL-it-Gemini-Pro-Heretic-Uncensored-Thinking_GGUF:Q4_K_M
# Run inference directly in the terminal:
./build/bin/llama-cli -hf Andycurrent/Gemma-3-4B-VL-it-Gemini-Pro-Heretic-Uncensored-Thinking_GGUF:Q4_K_M

Use Docker

docker model run hf.co/Andycurrent/Gemma-3-4B-VL-it-Gemini-Pro-Heretic-Uncensored-Thinking_GGUF:Q4_K_M

LM Studio
Jan

vLLM

How to use Andycurrent/Gemma-3-4B-VL-it-Gemini-Pro-Heretic-Uncensored-Thinking_GGUF with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "Andycurrent/Gemma-3-4B-VL-it-Gemini-Pro-Heretic-Uncensored-Thinking_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": "Andycurrent/Gemma-3-4B-VL-it-Gemini-Pro-Heretic-Uncensored-Thinking_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/Andycurrent/Gemma-3-4B-VL-it-Gemini-Pro-Heretic-Uncensored-Thinking_GGUF:Q4_K_M

Ollama
How to use Andycurrent/Gemma-3-4B-VL-it-Gemini-Pro-Heretic-Uncensored-Thinking_GGUF with Ollama:
```
ollama run hf.co/Andycurrent/Gemma-3-4B-VL-it-Gemini-Pro-Heretic-Uncensored-Thinking_GGUF:Q4_K_M
```

Unsloth Studio new

How to use Andycurrent/Gemma-3-4B-VL-it-Gemini-Pro-Heretic-Uncensored-Thinking_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 Andycurrent/Gemma-3-4B-VL-it-Gemini-Pro-Heretic-Uncensored-Thinking_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 Andycurrent/Gemma-3-4B-VL-it-Gemini-Pro-Heretic-Uncensored-Thinking_GGUF to start chatting

Using HuggingFace Spaces for Unsloth

# No setup required
# Open https://huggingface.co/spaces/unsloth/studio in your browser
# Search for Andycurrent/Gemma-3-4B-VL-it-Gemini-Pro-Heretic-Uncensored-Thinking_GGUF to start chatting

Docker Model Runner
How to use Andycurrent/Gemma-3-4B-VL-it-Gemini-Pro-Heretic-Uncensored-Thinking_GGUF with Docker Model Runner:
```
docker model run hf.co/Andycurrent/Gemma-3-4B-VL-it-Gemini-Pro-Heretic-Uncensored-Thinking_GGUF:Q4_K_M
```

Lemonade

How to use Andycurrent/Gemma-3-4B-VL-it-Gemini-Pro-Heretic-Uncensored-Thinking_GGUF with Lemonade:

Pull the model

# Download Lemonade from https://lemonade-server.ai/
lemonade pull Andycurrent/Gemma-3-4B-VL-it-Gemini-Pro-Heretic-Uncensored-Thinking_GGUF:Q4_K_M

Run and chat with the model

lemonade run user.Gemma-3-4B-VL-it-Gemini-Pro-Heretic-Uncensored-Thinking_GGUF-Q4_K_M

List all available models

lemonade list

Gemma-3-4B-VL-it-Gemini-Pro-Heretic-Uncensored-Thinking-GGUF

This repository contains Gemma-3-4B-VL-it-Gemini-Pro-Heretic-Uncensored-Thinking-GGUF, a 4B-parameter vision-language instruction-tuned model provided in GGUF format for efficient local inference. The model is designed for open-ended reasoning, multimodal understanding, and minimal alignment constraints, making it suitable for experimentation, research, and advanced local deployments.

Model Summary

Model ID: Gemma-3-4B-VL-it-Gemini-Pro-Heretic-Uncensored-Thinking-GGUF
Architecture: Gemma 3 (4B parameters)
Type: Vision-Language (Text + Image)
Format: GGUF
Publisher: mradermacher
License: Apache 2.0 (inherits from base model)

Key Characteristics

Multimodal input support (text + images)
Instruction-tuned for conversational and reasoning tasks
Reduced content filtering and alignment constraints
Optimized for local inference runtimes
Suitable for research, exploration, and advanced user workflows

⚠️ This model is uncensored. Outputs may include sensitive or unfiltered content. Use responsibly.

Supported Use Cases

Text-Based

Conversational assistants
Creative writing and storytelling
Summarization and rewriting
General reasoning and analysis

Vision + Text

Image captioning
Visual question answering
Scene and object understanding
Multimodal reasoning tasks

GGUF Compatibility

This model can be used with GGUF-compatible runtimes such as:

llama.cpp
Ollama (GGUF-based builds)
Other local inference engines supporting GGUF

Performance and supported features may vary depending on runtime and hardware.

Basic Usage Example

Command Line (llama.cpp-style)

./main \
  -m Andycurrent/Gemma-3-4B-VL-it-Gemini-Pro-Heretic-Uncensored-Thinking_GGUF_F16.gguf \
  -p "Describe the key idea behind multimodal AI models."