This is more complicated than some corporate infrastructures I’ve worked on, lol.

Yeah, the image bytes are random because they’re already compressed (unless they’re bitmaps, which is not likely).

OSMC’s Vero V looks interesting. Pi 4 with OSMC or Librelec could work. I’m probably going to do something like this pretty soon. I just set up an *arr stack last week, and just using my smart TV with the jellyfin app installed ATM.

My PC running the Jellyfin server can’t transcode some videos though; probably going to put an Arc a310 in it.

I like the Turris Omnia and (highly configurable) Turris Mox. They come with OpenWrt installed.

IDK, looks like 48GB cloud pricing would be 0.35/hr => $255/month. Used 3090s go for $700. Two 3090s would give you 48GB of VRAM, and cost $1400 (I’m assuming you can do “model-parallel” will Llama; never tried running an LLM, but it should be possible and work well). So, the break-even point would be <6 months. Hmm, but if Severless works well, that could be pretty cheap. Would probably take a few minutes to process and load a ~48GB model every cold start though?

I would if the hardware was powerful enough to do interesting or useful things, and there was software that did interesting or useful things. Like, I’d rather run an AI model to remove backgrounds from images or upscale locally, than to send images to Adobe servers (this is just an example, I don’t use Adobe products and don’t know if this is what Adobe does). I’d also rather do OCR locally and quickly than send it to a server. Same with translations. There are a lot of use-cases for “AI” models.