---
title: Qwen 7B Hosted Chat
emoji: 🤖
colorFrom: blue
colorTo: indigo
sdk: docker
app_port: 7860
---

# Qwen 2.5 7B - 4-bit Quantized

This Space runs a fully hosted Qwen 2.5 7B model using `llama.cpp` for CPU-friendly 4-bit quantization.

## Features
- **UI:** Gradio ChatInterface with session memory.
- **API:** FastAPI endpoint available at `/api/chat`.
- **Memory Footprint:** ~4.3GB (Fits in Free Tier).

## API Usage
You can send POST requests to `/api/chat`.

```python
import requests

url = "https://YOUR_SPACE_URL.hf.space/api/chat"
payload = {
    "messages": [
        {"role": "user", "content": "Hello!"}
    ]
}
response = requests.post(url, json=payload)
print(response.json())