Wanmory · Where Time Stays Warm
Bring your scattered photos together, organized by person, date, and location. Reconnect with memories and the places you’ve been, rediscover moments from this day in past years, and see how life has grown and changed over time.
Private photos, kept safe
Photos, thumbnails and the database all stay on your own hardware. No cloud services, no subscriptions. Your memories are yours to keep.
Local AI, your choice of compute
Face recognition, smart search and aesthetic scoring all run locally, and every model can be swapped. If your server is short on power, hand the work to a Mac or a GPU machine at home.
Organized automatically, time made visible
Photos are filed by date and your travels are traced on a map. On This Day brings old moments back, and comparing years shows how people and life have changed.
Shared with family, a space for everyone
Everyone gets their own account. Members manage the photos they upload, admins oversee everything, and the sign-in history is always available.
Six Core Features · A Fresh Way to Organize Your Photos
Slideshows, growth over the years, natural-language search, photo sharing, a travel map and burst cleanup, all on your own machine.
Slideshows1/6
Play any library, album, person or memory as a slideshow, with background music and transitions.
On-Device AI, Compute Where You Need It
A single low-power box is enough to get started, such as a NAS or a Raspberry Pi. Both amd64 and arm64 are supported.
The Docker setup ships with a built-in AI service that runs on the CPU alone. Even a low-power NAS such as an N100 or N5105 gets through your library at its own pace. Images are published for both amd64 and arm64.
Run the AI service directly on a Mac (outside a container) to use CoreML and the Apple GPU, while the NAS just stores the photos.
Run the AI service on a machine with a graphics card. ONNX Runtime uses CUDA, so you can switch to larger, more accurate face and search models.
Enter a remote address under Settings → AI service and AI jobs are handed to a more powerful machine on your network. Worker nodes can also pull jobs on their own.
Quick Install
Pull the images and install on any machine that runs Docker.
- 1Run the install script
Run one command on a machine with Docker installed. The script first checks for a 64-bit system, Docker, and enough memory and disk space.
$ curl -fsSL https://wanmory.com/install.sh | bash - 2Answer the prompts
Confirm the web port, LAN domain, database password (randomly generated by default) and its location, time zone (type a city to search), and data directory, then add one or more existing photo folders. The script writes the .env file for you.
- 3Starts automatically
The script downloads the official docker-compose.yml, pulls the images and starts everything. The database is initialized on first start.
- 4Open your library
Open http://wanmory.local:8080 in a browser, create the admin account, and download the models under Settings → AI.