Most downloaders fetch the file onto their own machine first and then pass it to you. This one does not, and the difference is visible.
Video arrives in pieces
Bluesky publishes video the way streaming services do: as a playlist pointing at a run of short pieces, each a few seconds long. There is no single finished MP4 sitting on their servers to link to.
So something has to fetch every piece and join them. The question is only whether that happens on a server in the middle or on your own device.
What changes when it happens on your device
There is no queue, because there is no shared machine to queue for. There is no size cap, because nobody is paying for the bandwidth but you and Bluesky. And there is no copy of your video sitting on someone else's disk afterwards.
The cost is the wait: your browser is doing real work, and on a slow connection a long video takes a while. The row of boxes filling in is that work made visible.
Checking it for yourself
Open your browser's developer tools before you paste a link, and watch the network panel. The requests for post details go to bsky.app. The requests for video go to Bluesky's video CDN. None of them carry video to or from this site.
This is the sort of claim that is easy to make and hard to prove, which is why it is worth showing you where to look rather than asking you to take it on trust.
What that means for images
Images work differently, and the honest answer is that they do pass through this server. Bluesky's image CDN refuses requests from other sites, so a browser cannot fetch an image and give it a sensible name on its own.
Rather than hand you a file called bafkreiev5vdn5i6c4rf and leave you to rename it, the image is streamed through this server and arrives named after the post. It is passed straight through and never stored.