PUT YOUR WARDROBE IN ROWS

The AI correctly identifies clothing type, color, pattern, and style in the vast majority of cases. It works best with clear, well-lit photos of individual items. You can always edit tags manually if needed — and the system learns from your corrections over time.

Wardrowbe works with any OpenAI-compatible API. Popular choices include Ollama (local, free), OpenAI (GPT-4o), and any LLM server that exposes a compatible endpoint. The cloud plan includes AI tagging out of the box — no configuration needed.

The cloud plan is instant — sign up and start adding clothes. For self-hosting, Docker Compose setup takes about 5 minutes if you already have Docker installed. The README walks you through every step.

Currently, migration is manual — you add items by photographing them. AI tagging makes this fast: snap a photo and the item is automatically categorized. Most users digitize their entire wardrobe in a single afternoon.

Completely. The self-hosted version runs entirely on your own server — no data ever leaves your network. The cloud plan stores data on our secure infrastructure, and we never share, sell, or use your data for training.

Any machine that can run Docker — a Raspberry Pi 5, an old laptop, a NAS, or a VPS. The app itself is lightweight. If you're running a local LLM (like Ollama), you'll want at least 8GB RAM and a decent CPU or GPU for fast tagging.