On this page
In short
Cloud AI video editors need the whole wedding uploaded first, and 100 to 500 GB can take many hours or even days on a home connection. Local tools skip the upload and keep the video on your computer, but need a capable PC. A cloud tool can suit a low-powered laptop or remote collaboration.
Every AI video editor needs a powerful computer to analyse your footage. A cloud tool uses its own servers, so the footage has to travel to them. A local tool uses your PC, so the footage stays where it is. When you weigh up local vs cloud video editing, that one difference decides how long you wait, what happens to your couples’ footage and what hardware you need. This article is part of our guide to AI wedding video editing.
What’s the difference between local and cloud AI editing?
A cloud editor runs in a web browser or a small app. You upload the wedding, the analysis and editing happen on the provider’s servers, and you review the result online or download it.
A local editor installs on your computer. It reads the cards, analyses the footage and renders the film on your own machine. Some local tools still use an online AI model for the planning stage. They send it still frames and a transcript while the video files stay on your computer.
How long does it take to upload a wedding?
Cloud AI video tools need the whole wedding before they can start, and that’s often 100 to 500 GB. Home broadband is often asymmetric, with an upload speed well below the download speed, so run a speed test and look at the upload figure.
To estimate the wait, use this formula:
Upload hours = GB × 8,000 ÷ Mbps ÷ 3,600
Multiplying by 8,000 turns gigabytes into megabits (8 bits in a byte, 1,000 megabytes in a gigabyte). Dividing by your upload speed in megabits per second gives seconds, and dividing by 3,600 turns seconds into hours.
Here’s what that gives for common wedding sizes and upload speeds.
| Wedding size | 20 Mbps | 50 Mbps | 100 Mbps | 500 Mbps |
|---|---|---|---|---|
| 100 GB | 11 h 7 min | 4 h 27 min | 2 h 13 min | 27 min |
| 300 GB | 33 h 20 min | 13 h 20 min | 6 h 40 min | 1 h 20 min |
| 500 GB | 55 h 33 min | 22 h 13 min | 11 h 7 min | 2 h 13 min |
These are best cases. The formula assumes your connection runs at full speed for the whole upload with nothing else using it, which real connections rarely manage. Your computer also has to stay on and connected until the last file arrives.
Try your own numbers here, or use the full upload time calculator.
Uploading to a cloud editor
13 h 20 min
Start it at 10 pm after the wedding and it finishes at 11:20 am the next day.
AfterRushes
Nothing to upload
Your footage stays on your computer. Only still frames and transcripts go to the AI.
Hours = GB × 8,000 ÷ Mbps ÷ 3,600, at full speed the whole time. Real uploads are often slower.
A long upload has knock-on effects. Your connection is busy while it runs, which can slow video calls and gallery uploads. If you promised the couple a quick teaser, the upload eats into that time before any editing starts. Sending a teaser days after the wedding covers that promise in more detail. And the upload comes on top of the backups you already make, which from camera card to backup walks through.
What happens to couples’ footage in each case?
With a cloud editor, the full wedding, pictures and sound, sits on someone else’s servers for as long as their policy allows. Before you use one, find out where the servers are, how long footage is kept, who can access it, and whether it’s used to train AI models.
With a local editor, the video files stay on your computer. Check what, if anything, the tool sends elsewhere. A local tool that uses an online AI model will send something to that model, such as still frames and a transcript, and you should know exactly what and where. Keeping couples’ footage private when you use AI lists the questions to ask, and our privacy page sets out exactly what AfterRushes sends.
What do you need to run AI editing locally?
Local editing asks more of your computer. Analysing hours of footage, transcribing speech and rendering films all take processing power, so a capable PC matters, and an NVIDIA graphics card with GPU acceleration makes the work faster.
On our test wedding, a 26-clip reception with 15 minutes of footage, AfterRushes went from camera card to finished preview films, AI included, in about 7 minutes on a PC with an NVIDIA graphics card using GPU acceleration. On a 2017 Intel i7 without GPU acceleration, the local analysis alone took about 11 minutes, including a one-time download, and the AI steps added a few minutes more. The time grows with the amount of footage, and the work runs in the background while you get on with other things.
You’ll also need enough drive space for the footage, which you’d need anyway to back it up. Check the requirements on the download page before you buy, and see video kit for hybrid wedding photographers if you’re planning an upgrade.
When might a cloud tool suit you better?
Local isn’t the right answer for everyone. A cloud tool may suit you if:
- you edit on a low-powered laptop and don’t plan to replace it
- you work with a remote editor or partner who needs to open the same project from somewhere else
- you have a fast upload, towards the right-hand column of the table, and your contract with the couple covers storing their footage with a third party
- you’d rather not own or maintain a powerful computer at all
If none of those apply, the upload time and the privacy question usually favour working locally. Either way, try any tool on a real wedding before you commit.