Posts on SelfVPS Guide

Deploying Open-Source AI Tools: LocalAI, Ollama, Stable Diffusion & More on Your VPS

Sat, 16 May 2026 14:00:00 +0800

Why Self-Host AI Tools?

As AI technology evolves rapidly, more open-source AI tools can run on your own server. Benefits include:

🔒 Data Privacy: Sensitive data never leaves your server
💰 Cost Control: Pay only for your hardware, no API subscription fees
⚡ Low Latency: Local inference, no network delays
🎯 Full Customization: Choose models and parameters freely

Hardware Requirements

Self-hosting AI tools requires some hardware. Recommended VPS specs:

Purpose	Minimum	Recommended
LLM (7B model)	8GB RAM, 4 vCPU	16GB RAM, 8 vCPU + GPU
Speech-to-Text	4GB RAM, 2 vCPU	8GB RAM, 4 vCPU
Image Generation	8GB RAM + 4GB VRAM	16GB RAM + 8GB VRAM

⚠️ Note: For GPU acceleration, consider providers like Hetzner (GPU cloud instances), RunPod, or Vast.ai.

Tool 1: Ollama — Run LLMs Locally

Ollama is the simplest way to run large language models locally. Supports Llama, Mistral, Qwen, and more.

Installation

# One-command Docker deploy
docker run -d --name ollama -p 11434:11434 \
 -v ollama:/root/.ollama \
 ollama/ollama

# Pull and run a model
docker exec -it ollama ollama pull llama3.2:1b
docker exec -it ollama ollama run llama3.2:1b

# API call
curl http://localhost:11434/api/generate -d '{
 "model": "llama3.2:1b",
 "prompt": "What is self-hosting?",
 "stream": false
}'

Recommended Models

Model	Parameters	RAM	Use Case
llama3.2:1b	1B	<2GB	Lightweight Q&A
llama3.2:3b	3B	~3GB	General chat
qwen2.5:7b	7B	~8GB	Chinese optimized
mistral:7b	7B	~8GB	English reasoning

Tool 2: LocalAI — OpenAI API Alternative

LocalAI is a drop-in replacement for OpenAI’s API, supporting LLM, TTS, image generation, and more.

# Docker Compose deploy
mkdir -p ~/localai && cd ~/localai

cat > docker-compose.yml << 'EOF'
version: '3.8'
services:
 localai:
 image: localai/localai:latest
 ports:
 - "8080:8080"
 volumes:
 - ./models:/build/models
 environment:
 - THREADS=4
 - CONTEXT_SIZE=2048
 command: ["/usr/bin/local-ai"]
EOF

docker compose up -d

Usage

# Chat completion (OpenAI SDK compatible)
curl http://localhost:8080/v1/chat/completions -d '{
 "model": "llama3.2-3b",
 "messages": [{"role": "user", "content": "Hello!"}]
}'

Tool 3: OpenAI Whisper — Speech-to-Text

Whisper is an open-source speech recognition model supporting 99+ languages.

# Docker deploy
docker run -d --name whisper \
 -p 9000:9000 \
 -v whisper-data:/data \
 onerahmet/openai-whisper-asr-webservice:latest

Use Cases:

Meeting transcription
Auto-generated video captions
Voice input systems

Tool 4: Stable Diffusion — Image Generation

Deploy via Automatic1111 WebUI:

# Docker deploy (requires GPU)
docker run -d --name sd-webui \
 --gpus all \
 -p 7860:7860 \
 -v models:/app/stable-diffusion-webui/models \
 abdibrokhim/stable-diffusion-webui:latest

Tool 5: LobeChat — AI Chat Interface

LobeChat is a modern AI chat UI supporting Ollama, LocalAI, and more.

# Docker deploy
docker run -d --name lobe-chat \
 -p 3210:3210 \
 -e OLLAMA_PROXY_URL=http://localhost:11434 \
 lobehub/lobe-chat:latest

Stack Architecture

Recommended self-hosted AI stack:

User → Nginx → LobeChat (frontend)
 ├── Ollama (LLM inference)
 ├── LocalAI (OpenAI-compatible API)
 └── Whisper (speech recognition)

Summary

In 2025, self-hosting AI tools has evolved from a niche experiment to a practical solution. With falling hardware costs and improving model optimization, running AI on your personal VPS is more accessible than ever.

Quick Start

Start with Ollama + LobeChat for the simplest setup
Add Whisper for speech processing as needed
Add Stable Diffusion when GPU is available

💡 Tip: If your VPS has limited resources, start with 1B-3B parameter models and scale up gradually.

VPS Cost-Saving Guide: Best Value Cloud Servers in 2025

Sat, 16 May 2026 12:00:00 +0800

Introduction

Cloud computing costs are on the rise. Whether you’re a solo developer, freelancer, or small startup, choosing a cost-effective VPS can save you significant money. This guide compares major cloud providers on price and performance.

Major VPS Provider Comparison

1. Hetzner Cloud ⭐ Best Value

Specs	Price	Use Case
2 vCPU / 4GB / 40GB	€4.15/mo	Small apps / n8n
4 vCPU / 8GB / 80GB	€8.85/mo	Medium apps / GitLab
8 vCPU / 16GB / 160GB	€18.35/mo	Large apps / Databases

Pros: Extremely low prices, great performance, EU & NA datacenters Cons: International focus, limited Chinese support

2. DigitalOcean

Specs	Price	Use Case
1 vCPU / 1GB / 25GB	$6/mo	Personal blog
2 vCPU / 2GB / 60GB	$12/mo	Small apps
2 vCPU / 4GB / 80GB	$24/mo	Medium apps

Pros: Excellent docs, active community, clean UI Cons: Relatively expensive, limited bandwidth

3. Vultr

Specs	Price	Use Case
1 vCPU / 1GB / 25GB	$6/mo	Personal projects
2 vCPU / 2GB / 55GB	$12/mo	Small apps
2 vCPU / 4GB / 80GB	$24/mo	Medium apps

Pros: Global locations, hourly billing Cons: Average performance, slow support

4. AWS Lightsail

Specs	Price	Use Case
1 vCPU / 512MB / 20GB	$3.5/mo	Dev/test
1 vCPU / 1GB / 40GB	$5/mo	Small sites
2 vCPU / 2GB / 60GB	$10/mo	Small apps

Pros: AWS ecosystem integration, transparent pricing Cons: Lower performance than dedicated VPS, fewer options

Money-Saving Tips

1. Choose the Right Billing Cycle

Most providers offer monthly or yearly discounts:

Hetzner: Hourly billing, flexible
Vultr: Hourly billing, start/stop anytime
DigitalOcean: Monthly, occasional annual deals

2. Use Free Credits

Google Cloud: $300 free credit for new users (90 days)
Oracle Cloud: Always-free tier (1 vCPU/1GB ARM)
AWS: 12-month free tier (750 hours t2.micro/month)

3. Monitor Resource Usage

Don’t pay for a high-end VPS if you only use a fraction. Use htop, df -h, and other tools to monitor usage and adjust accordingly.

My Recommendations

Budget Priority: Hetzner Cloud — €4.15/mo for 2 vCPU / 4GB Beginner Friendly: DigitalOcean — great docs and community Asia Optimized: Alibaba Cloud or Hong Kong node VPS

Summary

When choosing a VPS, look beyond price. Consider performance, network quality, datacenter location, and support. Start with a low-cost plan and scale up as needed.

💡 Tip: Many providers offer free trials. Test actual network latency and performance before committing.

N8N Deployment Guide: Self-Host Workflow Automation on a VPS with Docker

Sat, 16 May 2026 10:00:00 +0800

Why N8N?

N8N is an open-source workflow automation tool, similar to Zapier and Make, but you deploy it on your own server for full data control and privacy.

N8N Advantages

🔓 Open Source & Free: Community edition is completely free with no user limits
🔒 Self-Hosted Data: Sensitive data never leaves your server
🔌 400+ Integrations: Slack, GitHub, Gmail, Discord, and more
🧩 Visual Builder: Drag-and-drop editor, no coding required
🏢 Extensible: Custom nodes and scripts supported

Prerequisites

A VPS (2GB RAM recommended)
Docker and Docker Compose installed
A domain name (optional, for HTTPS)
Basic Linux command line knowledge

Step 1: Install Docker

# Install Docker
curl -fsSL https://get.docker.com | sh

# Enable Docker on boot
sudo systemctl enable docker
sudo systemctl start docker

# Install Docker Compose plugin
sudo apt-get install docker-compose-plugin -y

Step 2: Create Docker Compose Configuration

mkdir -p ~/n8n && cd ~/n8n

Create docker-compose.yml:

version: '3.8'

services:
 n8n:
 image: n8nio/n8n:latest
 container_name: n8n
 restart: unless-stopped
 ports:
 - "5678:5678"
 environment:
 - N8N_HOST=${N8N_HOST}
 - N8N_PORT=5678
 - N8N_PROTOCOL=https
 - WEBHOOK_URL=https://${N8N_HOST}/
 - DB_TYPE=postgresdb
 - DB_POSTGRESDB_HOST=postgres
 - DB_POSTGRESDB_DATABASE=${POSTGRES_DB}
 - DB_POSTGRESDB_USER=${POSTGRES_USER}
 - DB_POSTGRESDB_PASSWORD=${POSTGRES_PASSWORD}
 - N8N_ENCRYPTION_KEY=${N8N_ENCRYPTION_KEY}
 volumes:
 - n8n_data:/home/node/.n8n
 depends_on:
 postgres:
 condition: service_healthy
 networks:
 - n8n_network

 postgres:
 image: postgres:15-alpine
 container_name: n8n-postgres
 restart: unless-stopped
 environment:
 - POSTGRES_USER=${POSTGRES_USER}
 - POSTGRES_PASSWORD=${POSTGRES_PASSWORD}
 - POSTGRES_DB=${POSTGRES_DB}
 volumes:
 - postgres_data:/var/lib/postgresql/data
 healthcheck:
 test: ["CMD-SHELL", "pg_isready -U ${POSTGRES_USER} -d ${POSTGRES_DB}"]
 interval: 10s
 timeout: 5s
 retries: 5
 networks:
 - n8n_network

volumes:
 n8n_data:
 postgres_data:

networks:
 n8n_network:
 driver: bridge

Step 3: Configure Environment Variables

Create .env file:

cat > .env << 'EOF'
N8N_HOST=n8n.yourdomain.com
N8N_ENCRYPTION_KEY=$(openssl rand -hex 32)
POSTGRES_USER=n8n
POSTGRES_PASSWORD=$(openssl rand -base64 32)
POSTGRES_DB=n8n
EOF

⚠️ Security Note: Always replace N8N_ENCRYPTION_KEY and POSTGRES_PASSWORD with strong, unique values.

Step 4: Configure Nginx Reverse Proxy

server {
 listen 80;
 server_name n8n.yourdomain.com;
 return 301 https://$server_name$request_uri;
}

server {
 listen 443 ssl http2;
 server_name n8n.yourdomain.com;

 ssl_certificate /etc/letsencrypt/live/n8n.yourdomain.com/fullchain.pem;
 ssl_certificate_key /etc/letsencrypt/live/n8n.yourdomain.com/privkey.pem;

 location / {
 proxy_pass http://localhost:5678;
 proxy_http_version 1.1;
 proxy_set_header Upgrade $http_upgrade;
 proxy_set_header Connection "upgrade";
 proxy_set_header Host $host;
 proxy_set_header X-Real-IP $remote_addr;
 proxy_set_header X-Forwarded-Proto $scheme;
 proxy_set_header X-Forwarded-For $proxy_add_x_forwarded_for;
 }
}

Step 5: Get SSL Certificate

# Install acme.sh
curl https://get.acme.sh | sh

# Issue certificate
acme.sh --issue -d n8n.yourdomain.com --nginx

# Install certificate
acme.sh --install-cert -d n8n.yourdomain.com \
 --key-file /etc/nginx/ssl/n8n.key \
 --fullchain-file /etc/nginx/ssl/n8n.crt

Step 6: Launch N8N

cd ~/n8n
docker compose up -d

Check the logs:

docker compose logs -f n8n

Visit https://n8n.yourdomain.com to complete the initial setup.

Management Commands

# View logs
docker compose logs -f n8n

# Restart service
docker compose restart n8n

# Update N8N
docker compose pull n8n
docker compose up -d n8n

# Backup database
docker exec n8n-postgres pg_dump -U n8n n8n > backup_$(date +%Y%m%d).sql

Summary

With Docker Compose, you can deploy n8n in under 10 minutes. Combined with PostgreSQL, Nginx, and Let’s Encrypt, you’ll have a production-grade workflow automation platform running on your own VPS.

Next, explore n8n’s 400+ integration nodes and start building your own automated workflows!