Repolode
Submit
kossakovsky

kossakovsky/selfhost-ai

One-command Docker Compose installer that deploys a self-hosted AI and automation stack (n8n, Ollama, Open WebUI, Dify, Supabase and 30+ tools) on your own VPS.

Open-source alternative to:ZapierMake

View on GitHubWebsite
Stars
951
Forks
239
Open issues
0
Last commit
21 hours ago

What is it?

Selfhost AI is a Docker Compose installer for a private AI homelab on a VPS running Ubuntu 24.04. An interactive wizard asks for a domain and email, lets you pick from more than 30 services, generates all secrets, and places the chosen services behind Caddy with automatic Let's Encrypt HTTPS. Available services include n8n for workflow automation, Ollama for local LLMs, Open WebUI as a chat interface, Dify, Flowise and Letta for agents, Qdrant and Weaviate for vector search, Supabase, ComfyUI, and Grafana with Prometheus for monitoring.

The installer also configures the firewall and brute-force protection, installs Docker, and can optionally import a set of community n8n workflows. Updates are run with make update, and custom Caddy sites and Compose overrides are kept across upgrades. Telemetry is sent through Scarf and can be disabled in the .env file.

The project requires a public VPS with a wildcard DNS record; home servers and localhost setups are not supported. Recommended hardware ranges from 4 GB RAM for n8n and monitoring alone to 20 GB RAM and 60 GB disk for the full stack.

Key features

  • 01Interactive wizard selects services and generates all secrets automatically
  • 02Automatic HTTPS for each service through Caddy and Let's Encrypt
  • 03Local LLMs via Ollama on CPU, NVIDIA or AMD GPUs, including multi-GPU
  • 04n8n workflow automation with queue mode and optional community workflows
  • 05Grafana and Prometheus dashboards and alerts for system and n8n metrics
  • 06make update, make doctor and other commands for upgrades and diagnostics

Use cases

  • Running a private ChatGPT-style interface and local models on your own server
  • Replacing hosted automation services like Zapier or Make with self-hosted n8n
  • Building RAG applications with Qdrant, Weaviate or Dify on your own infrastructure
  • Setting up a personal homelab with AI agents and monitoring on a VPS

Quick start

  1. 1

    Point a wildcard DNS record (*.yourdomain.com) at your server's IP

  2. 2

    Clone the repo and run the installer

    git clone https://github.com/kossakovsky/selfhost-ai && cd selfhost-ai && sudo bash ./scripts/install.sh
  3. 3

    Follow the interactive wizard to choose services and enter your domain and email

Things to know

  • Needs a VPS with a public IP and a domain; home servers and localhost setups are not supported.
  • Designed for Ubuntu 24.04 LTS; other operating systems are not covered in the README.
  • Running all services requires at least 20 GB RAM, 4 CPU cores and 60 GB disk.

Topics

#self-hosted#docker-compose#llm#ollama#workflow-automation#n8n#rag#homelab#ai-agents#ai-automation

FAQ

What is selfhost-ai?
One-command Docker Compose installer that deploys a self-hosted AI and automation stack (n8n, Ollama, Open WebUI, Dify, Supabase and 30+ tools) on your own VPS. Selfhost AI is a Docker Compose installer for a private AI homelab on a VPS running Ubuntu 24.04. An interactive wizard asks for a domain and email, lets you pick from more than 30 services, generates all secrets, and places the chosen services behind Caddy with automatic Let's Encrypt HTTPS. Available services include n8n for workflow automation, Ollama for local LLMs, Open WebUI as a chat interface, Dify, Flowise and Letta for agents, Qdrant and Weaviate for vector search, Supabase, ComfyUI, and Grafana with Prometheus for monitoring.
Is selfhost-ai free to use? What license does it have?
selfhost-ai is open source under the Apache-2.0 license. Check the license terms before using it in commercial products.
Is selfhost-ai still maintained?
Actively maintained. The last commit to selfhost-ai was on Oct 19, 2026.
What can I use selfhost-ai for?
Running a private ChatGPT-style interface and local models on your own server Replacing hosted automation services like Zapier or Make with self-hosted n8n Building RAG applications with Qdrant, Weaviate or Dify on your own infrastructure Setting up a personal homelab with AI agents and monitoring on a VPS
How do I install selfhost-ai?
Point a wildcard DNS record (*.yourdomain.com) at your server's IP Clone the repo and run the installer: git clone https://github.com/kossakovsky/selfhost-ai && cd selfhost-ai && sudo bash ./scripts/install.sh Follow the interactive wizard to choose services and enter your domain and email
Where can I get selfhost-ai?
The source code and installation instructions are on GitHub: https://github.com/kossakovsky/selfhost-ai

Summary generated by AI from the project's README. GitHub stats refresh daily.

Related projects