How TwineLine works
TwineLine puts together three pieces to create and play dynamic slideshows:
Photos & videos
Photos and videos in a local drive and/or online collections like OneDrive, Google Photos, iCloud, Pixabay, etc.
TwineLine server
AI organizes your photos by people, places, and moments to create a unique slideshow each day.
Devices
Slideshows can be viewed on almost any device with a modern browser, all at the same time.
TwineLine server
The TwineLine server is the “brains” of the app. It analyzes photos, creates slideshows, synchronizes playback on multiple devices, and responds to user controls.
Your biggest choice (which you can change at any time) is where the server runs. You can run a local server on a computer in your home for free (Windows, Mac, or Linux). This provides maximum control and privacy.
Alternatively, a cloud server can run for less than $7 per month. This is the easiest option, provides good speed for AI analysis, and allows your friends and family to see your daily slideshow (protected by a password) anywhere in the world. You can try it for free for a month and then move to your own computer if you prefer.
If you’re torn between these options, we suggest trying the cloud server. It can do a lot of AI analysis of your collection during the free month. The results can be copied to your own computer later if you prefer not to pay for the cloud server (and don’t care about web access for your slideshows).
Photos & videos
A local server can create slideshows from photos and videos stored on your computer’s hard drive. Both types of server can access photos on OneDrive as well as public photo collections like Pixabay.
Why OneDrive? Because Microsoft provides operating system independent access to your online photo collection that TwineLine needs. Apple and Google do not. But don’t worry — TwineLine can copy large collections from Google/iCloud Photos to OneDrive if you can wait a day or two for the task to complete.
If you don’t have a OneDrive account, you can get an account with 5 GB of storage for free. That’s only enough for a small photo collection, but an account with 1 TB of storage (1000 GB) costs less than $9 per month paid annually. More details.
Devices
A TwineLine slideshow can be viewed or controlled using any device with a modern browser (Chrome, Safari, and Edge have been tested). To connect to a TwineLine server on a local computer, simply enter the computer’s IP address in the browser’s search bar (the device and the server must be on the same local network).
Focus on flexibility
TwineLine is very flexible about where your photos and videos are stored, what server computer you have or rent, and the kinds of devices that can display your slideshows. It might take a little time to get things perfectly set up for your wishes and your budget.
Where are your files?
Different companies have different policies about how apps like TwineLine can access photos and videos in their collections.
Microsoft OneDrive provides the best policy, allowing TwineLine to directly access the files and location data recorded by phones and modern cameras.
Apple also provides good access, but only if you are running TwineLine on a Mac computer. Apple does not provide a good way to migrate your files to OneDrive, so a Mac is required.
Google Photos closed access in 2025, making things difficult but not impossible. TwineLine provides ways to migrate your Google Photos to OneDrive where everything works smoothly.
Smart TV + your computer
Best if your TV has Google TV or Fire TV built in. No extra hardware needed.
Your photos
Your computer
Smart TV
Cost: free. Runs when your computer is on.
Streaming stick + your computer
Best for older TVs without a built-in smart platform.
Your photos
Your computer
Stick + TV
HDMI
Cost: ~$30 one-time for the stick. Runs when your computer is on.
All-in-one computer
Best for a dedicated mini PC that stays connected to the TV.
Your photos
Computer
server + display
Your TV
Cost: free (if you have a spare computer) or ~$150 for a mini PC.
Cloud server
Best for always-on operation and access from anywhere.
Your photos
Hetzner cloud
~$7/mo
Any screen
TV, phone, tablet
Cost: ~$7/month (at least one month free with our referral link).
Which setup is right for me?
There is no wrong choice — you can change your setup anytime without losing your photos, face recognition data, or slideshows.
| Your computer | Cloud (Hetzner) | |
|---|---|---|
| Cost | Free forever | ~$7/month (at least 1 month free) |
| Always running? | Only when your computer is on | Yes, 24/7 |
| Access outside your home? | No (home network only) | Yes, from anywhere |
| Fast AI analysis | Standard speed | Optional turbo mode ($10–$20 one-time) |
| Best for | Trying it out, single household | Daily automated slideshows, families across multiple homes |
Our suggestion
Start with a Hetzner cloud server. With our referral link, you get at least one month of free hosting. You get 24/7 operation, web access from your phone, and the option to run fast AI analysis on your whole photo library. If you decide to switch to your own computer later, you can migrate everything with a single command and run TwineLine for free going forward.
Quick guide
If your TV has Google TV or Fire TV built in — you only need to set up the server. No extra hardware.
If your TV is an older model — add a streaming stick (~$30) that plugs into your TV.
If you want to dedicate a computer to TwineLine — connect it directly to your TV with an HDMI cable. It handles everything.
Want it always on without managing a computer? — use a cloud server. Free for the first couple of months with our referral link.
How it works technically
Server: Python FastAPI application serving a REST API on port 8080.
Player: Vue.js 3 single-page app optimized for streaming devices, rendered via the TwineLine app or any Chromium-based browser.
Communication: WebSocket connection with a 1-second heartbeat. The server is the single source of truth for slideshow state — slide index, play/pause, timing. Clients are passive viewers.
Database: SQLite (single-file, no setup required).
Photo sources: Microsoft OneDrive (via Microsoft Graph API), local filesystem, Pixabay stock photos.
Image processing: Pillow for optimization, OpenCV for face detection, FFmpeg for video transcoding.
Face recognition: FaceNet model running locally on CPU (~2 GB model download, no cloud dependency, no GPU required).
Multi-device sync: A lightweight signaling layer keeps all screens in sync. Only control messages travel through this layer — your photos never leave your network.