I’m building almost exactly this as an N150 box, so take this with that bias. For photos, files and contacts, Immich plus Nextcloud on Debian will cover you, and an N-series board idles at a few watts. The part that bit me was updates breaking things, not the install. What’s your limit for maintenance, an hour a month?
I’m designing around an Intel N150, and this thread is the reason. These chips idle in single digits with an SSD, so most of the cost of running always-on goes away. The catch is the 16GB RAM ceiling on most boards, which caps how much local AI you can run on it. For the low-watt folks: what did you give up to get there, and was it worth it?
The layout I’m designing around is inline: ISP router, then a small dual-ethernet box doing firewall and DNS filtering, then your LAN. For a start, I’d put OPNsense on a cheap dual-NIC mini PC. You get network-wide blocking through a plugin and a VPN to reach your NAS, and VLANs can wait until you need them. I’m biased, since I’m building a box with that exact layout. What I can’t settle: for someone starting out, is an all-in-one box less scary, or does it hide the stuff you’d want to learn?
Calling it ad blocking undersells it. The ad network is the delivery path in that report, so blocking at DNS is a security layer, not tidiness. The thing I’m building goes a step further: it holds tracker connections open instead of refusing them, so they waste their own time instead of retrying instantly. Has anyone tried tarpitting at home? I’m curious whether it broke smart TVs or apps that phone home hard.
Yes, but the gift isn’t the hardware. It’s that it keeps working when you’re not around to fix it. So I’d go boring: plain Debian, Cockpit for storage and containers, and a WireGuard web UI. Before you hand it over, unplug the internet for a day and see what breaks. DNS usually goes first. I’m building an open-source home box around the same idea (GPLv3 once it ships), and your question is the one I keep chewing on: what makes a partner actually use it instead of just tolerating it?
I’m building toward exactly this, so take it with that bias. The cheap tier works if you lower expectations. An N150 mini PC with 16GB runs small models through Ollama fine, but 8B models are slow on that class of chip, and you’ll feel it on anything chat-shaped. Where it shines is async work: summarize this, tag that, answer when it’s ready. For phone access I’d keep it LAN-only and use WireGuard or Tailscale to get home rather than opening a port.
Disclosure: I’m working on a small open-source home box that bundles local AI with offline Wikipedia and media. What would make you trust a box like that over a setup you built yourself?
Ollama on an N150 mini PC, 16GB, no GPU, Open WebUI on top. Being honest, 1-3B models are fine for searching my offline Kiwix stuff and 8B works if you don’t mind waiting. I’m turning this into a small box (heyeddi.com/projects/doomsday-box) and I keep going back and forth on whether people even want AI on it, or just the media and offline archives. Would you?