---
title: SuperGemma4-26B Uncensored
emoji: 🚀
colorFrom: purple
colorTo: pink
sdk: gradio
sdk_version: 6.16.0
app_file: app.py
pinned: false
license: apache-2.0
python_version: 3.11
---

# SuperGemma4-26B Uncensored (GGUF - CPU)

Powerful 26B parameter uncensored AI model running on CPU using GGUF format.

## Features

- 💬 Interactive chat without restrictions
- 💻 Code generation (Python, JavaScript, Java, C++, etc.)
- 📝 Uncensored text generation
- 🔥 26B parameters with Q4_K_M quantization
- ⚡ **Runs on CPU** - No GPU required!

## Model

- **Base**: Jiunsong/supergemma4-26b-uncensored-gguf-v2
- **Parameters**: 26 billion
- **Quantization**: Q4_K_M (4-bit)
- **Type**: Uncensored
- **Format**: GGUF (llama.cpp compatible)

## Hardware

This Space runs on **CPU** - no GPU required! Uses llama.cpp for efficient CPU inference.

## Performance

- **Speed**: ~2-5 tokens/second on CPU
- **Memory**: ~16GB RAM
- **Context**: 4K tokens

## Usage

### Web Interface

Use the tabs above to chat, generate code, or create text.

### API

```python
from gradio_client import Client

client = Client("sagarleet/test")
result = client.predict("Write a Python function", 500, 0.7, 0.9, 40, api_name="/predict")
print(result)
```

## Disclaimer

This is an uncensored AI model. Use responsibly and ethically.