AI LaunchKit: One-command installer for n8n + 50+ AI tools (Ollama, ComfyUI, RAG, STT/TTS, vector DBs)

Built with n8n
,
1

What It Is

AI LaunchKit deploys n8n alongside 50+ AI and automation tools with automatic integration via Docker Compose. All services communicate internally without API keys.

:link: GitHub: GitHub - freddy-schuetz/ai-launchkit: 🚀 AI LaunchKit - Complete self-hosted AI development toolkit with 50+ pre-configured tools including n8n, bolt.diy, ComfyUI, and more. One-command installation.
:scroll: License: Apache 2.0

Screenshot 2025-10-18 194319
Screenshot 2025-10-18 194319910×816 41.2 KB

n8n Configuration

  • Production mode with PostgreSQL backend
  • Queue mode enabled (parallel execution)
  • 300+ community workflows pre-imported
  • Media tools pre-installed (FFmpeg, ImageMagick, Python3)
  • Redis for queue management
  • Automatic HTTPS: https://n8n.yourdomain.com

Integrated AI Tools

Screenshot 2025-09-29 000423
Screenshot 2025-09-29 0004231146×900 95.9 KB

All accessible via HTTP Request nodes using internal Docker network:

:brain: LLMs

  • Ollama (http://ollama:11434) - Llama 3, Mistral, Gemma locally
  • Open WebUI - ChatGPT-like interface
  • Letta - Stateful agents with memory
  • Dify - LLM operations platform

:artist_palette: Image Generation

  • ComfyUI (http://comfyui:8188) - Stable Diffusion workflows

:books: RAG & Vectors

  • Qdrant (http://qdrant:6333) - Vector database
  • Weaviate - AI-native vector DB
  • RAGApp - Document Q&A system
  • LightRAG - Graph-based retrieval

:studio_microphone: Speech

  • Faster-Whisper (http://whisper:8001) - Speech-to-text (OpenAI compatible)
  • OpenedAI-Speech (http://openedai:5001) - Text-to-speech
  • Scriberr - Meeting transcription with speaker ID
  • Vexa - Real-time transcription API

:magnifying_glass_tilted_left: Search & Data

  • SearXNG (http://searxng:8080) - Privacy-respecting metasearch
  • Crawl4Ai - AI-optimized web scraping
  • GPT Researcher - Autonomous research agent
  • Perplexica - AI-powered search

:page_facing_up: Documents

  • Gotenberg (http://gotenberg:3000) - Universal PDF converter
  • Stirling-PDF - 100+ PDF operations
  • OCR Bundle - Tesseract + EasyOCR

:shield: Security

  • LLM Guard - Prompt injection detection
  • Presidio - PII detection (English)
  • Flair NER - PII detection (German)

:floppy_disk: Infrastructure

  • Supabase - Auth, storage, PostgreSQL + pgvector
  • PostgreSQL - n8n database
  • Redis - Queue backend
  • Neo4j - Knowledge graphs

:robot: Automation

  • Flowise - Visual AI agent builder
  • Browser-use - LLM browser control
  • Browserless - Headless Chrome

:briefcase: Business Tools (Optional)

Cal.com, Invoice Ninja, Vikunja, Baserow, NocoDB, Kimai, Odoo, Twenty CRM, EspoCRM, Mautic, and more.

Full list

Example Workflows

Local LLM Content Generation

Schedule → HTTP Request (Ollama) → Format → Supabase

Meeting Transcription

Webhook (audio) → Whisper STT → Ollama (summarize) → Email

AI Image Pipeline

Schedule → Ollama (prompt) → ComfyUI (generate) → Supabase → Social media

RAG Support Bot

Webhook (question) → Qdrant (search) → Ollama (answer) → Response

Research Automation

Trigger → SearXNG → Crawl4Ai → Ollama (analyze) → PDF Report

Document OCR

Email (PDF) → OCR → Ollama (extract data) → Supabase

Security Check

Input → LLM Guard → Presidio (PII) → Ollama → Sanitize

HTTP Request Example

Calling Ollama from n8n:

POST http://ollama:11434/api/generate

{
  "model": "llama3",
  "prompt": "{{$json.userQuestion}}",
  "stream": false
}

No API keys, no rate limits, internal network only.

Installation

git clone https://github.com/freddy-schuetz/ai-launchkit
cd ai-launchkit
sudo bash ./scripts/install.sh

Interactive wizard → select services → 10-30 minutes.

Screenshot 2025-08-28 203207
Screenshot 2025-08-28 203207952×573 16.3 KB

Hardware Requirements

  • Minimal (n8n + basic tools): 8GB RAM, 100GB disk
  • Comfortable (+ LLMs): 8-32GB RAM, 200GB disk
  • Full stack (+ GPU): 32GB+ RAM, 300GB+ disk

Key Features

:white_check_mark: All services on internal Docker network (no auth needed)
:white_check_mark: External HTTPS via Caddy (automatic certificates)
:white_check_mark: Profile-based deployment (install only what you need)
:white_check_mark: Update script included
:white_check_mark: Security hardened (firewall, fail2ban, auto-passwords)

Technical Details

  • Built on kossakovsky/n8n-installer
  • Docker Compose with profiles
  • Caddy reverse proxy
  • PostgreSQL + Redis backends
  • All integrations pre-configured

Common Questions

Q: Existing n8n setup?
A: Fresh installations only.

Q: Select specific tools?
A: Yes, wizard lets you choose.

Q: Updates?
A: sudo bash ./scripts/update.sh

Q: GPU for ComfyUI?
A: Requires NVIDIA Docker setup (documented).

Q: Import workflows?
A: Standard n8n import/export works.

Links

Feedback welcome!