en
Self-Hosted AI Tools on VPS - Private LLM and Generative AI
Sending sensitive business data to OpenAI, Anthropic, or Google's APIs means that data is processed on their servers, potentially used for training, and subject to US law and subpoenas. Self-hosted AI tools on a VPS keep your data entirely under your control. Ollama (LLM inference), Stable Diffusion (image generation), and Whisper (transcription) all run effectively on modern VPS hardware.
Need this done for your project?
We implement, you ship. Async, documented, done in days.
Ollama LLM Server on VPS
Ollama runs open-source LLMs locally. Install on your VPS:
```bash
curl -fsSL https://ollama.ai/install.sh | sh
# Pull a model (Llama 3 8B is a good starting point):
ollama pull llama3
# Run API server
ollama serve
```
API endpoint at http://YOUR_VPS_IP:11434. Compatible with OpenAI API format:
```python
import openai
client = openai.OpenAI(
base_url="http://YOUR_VPS_IP:11434/v1",
api_key="ollama" # arbitrary key
)
response = client.chat.completions.create(
model="llama3",
messages=[{"role": "user", "content": "Your sensitive business query here"}]
)
```
LLaMA 3 8B runs on 8GB RAM (quantized). 70B models require 48GB+ RAM. For business-quality outputs without GPU, LLaMA 3 8B-instruct is surprisingly capable.
Private Data Processing with Whisper
OpenAI's Whisper speech recognition model runs locally. Medical transcription, legal recordings, and confidential meeting transcription should not go to cloud APIs.
```bash
apt install python3-pip ffmpeg -y
pip install openai-whisper
whisper meeting_recording.mp3 --model medium --language en --output_format srt
```
The "medium" model is accurate and requires 5GB RAM. "large-v3" provides better accuracy with 10GB RAM.
For automated batch processing:
```bash
for audio in /input/*.mp3; do
whisper "$audio" --model medium --output_format txt --output_dir /output/
done
```
All audio stays on your VPS. No API calls to US-based services. Transcripts are stored only on your Iceland or Romania server.
Related Services
Why Anubiz Host
100% async — no calls, no meetings
Delivered in days, not weeks
Full documentation included
Production-grade from day one
Security-first approach
Post-delivery support included
Ready to get started?
Skip the research. Tell us what you need, and we'll scope it, implement it, and hand it back — fully documented and production-ready.