36

Does anyone run one of the above on a Pi 4 and can share their experience how good or bad they run?

If course, transcoding won't be any good and OCR probably cannot run in parallel, but aside from that - is it okay?

Currently running everything on a mini ITX with a i5-6600 which handles this easily for my small use cases, but also draws 20-30W idling most of the day... I'm eyeing a Pi 4b with 8gb RAM but don't want to spend the money and then realizing that it doesn't run smooth enough

top 16 comments
sorted by: hot top controversial new old
[-] mbp@slrpnk.net 6 points 2 days ago* (last edited 2 days ago)

I tried paperless years ago on my 4B and it did not work well enough to be usable.

Jellyfin was fine.

I'd say getting an x86 think centre or equivalent will cut your idle in half and give you enough overhead to run paperless and jellyfin.

[-] frosch@sh.itjust.works 1 points 2 days ago

it did not work well enough to be usable.

Was it the overall performance or the OCR specifically?

I have run paperless some time on a pi 3b without OCR (manually doing it on a PC or when scanning with Apps like MakeACopy) and it was okay-ish. Not a lot documents though.

And I thought it was mainly the 1GB RAM limiting (starting Paperless started swapping right away...)

[-] mbp@slrpnk.net 1 points 1 day ago

I remember it also having some issues when browsing the scanned files but I did use OCR as that was a requirement for paperless being useful in my case.

[-] Onomatopoeia@lemmy.cafe 2 points 2 days ago

I run a 2019 Dell OptiPlex SFF desktop as my ESXi box - it idles under 20w with multiple Linux and Windows VM's (4 are standard, besides the ad-hoc ones for testing stuff).

Hard to beat the idle combined with performance when needed. Pi really doesn't compare.

[-] broodmother@discuss.tchncs.de 2 points 2 days ago

Zu Paperless-ngx auf dem Pi 4: Für leichte private Dokumentenmengen kann das funktionieren, aber OCR ist der Teil, der schnell zäh wird. Ich würde die Worker klein halten und OCR/Tika/Gotenberg eher auslagern, wenn regelmäßig größere PDFs reinkommen. Paperless selbst ist meist weniger das Problem als OCR + I/O; Jellyfin parallel ist ohne Transcoding deutlich entspannter.

[-] frosch@sh.itjust.works 1 points 1 day ago

Ja, weniger Worker, dann dauert das OCR eben Mal länger. Da das eher selten ich muss das Dokument sofort im Paperless anschauen, kann das meinetwegen auch seine Minuten brauchen.

Bin bisher auch noch nicht 100% auf den Workflow eingeschossen:

  • MakeACopy macht ganz gutes OCR auf dem Handy, die Scans sind aber oft etwas wackelig.
  • FairScan macht sehr gute PDFs, aber kein OCR
  • Ich habe einen Scanner, der natürlich bei 600DPI nahezu perfekte Kopien macht, die ich auf dem PC per Shell in Sekunden durch OcrMyPDF jagen kann für extrem gute Ergebnisse. Aber>5 Seiten per Hand auf dem Scannerbett... das ist dann schon nervig
[-] KatherinaReichelt@feddit.org 1 points 1 day ago* (last edited 1 day ago)

Als Privatnutzer bekommt man ja eher seltener gigantische Mengen an Papier per Post. Das sind dann maximal ein paar Briefe pro Woche und dann ist es auch total egal, ob die OCR 10 Minuten pro Brief braucht. Und wenn du als Firma wirklich noch eine größere Menge an Briefpost zu verwalten hast, dann ist ein Raspi definitiv die falsche Wahl

[-] Decronym@lemmy.decronym.xyz 1 points 1 day ago* (last edited 1 day ago)

Acronyms, initialisms, abbreviations, contractions, and other phrases which expand to something larger, that I've seen in this thread:

Fewer Letters More Letters
AP WiFi Access Point
DHCP Dynamic Host Configuration Protocol, automates assignment of IPs when connecting to a network
DNS Domain Name Service/System
ESXi VMWare virtual machine hypervisor
IP Internet Protocol
MQTT Message Queue Telemetry Transport point-to-point networking
PiHole Network-wide ad-blocker (DNS sinkhole)
SSH Secure Shell for remote terminal access
SSL Secure Sockets Layer, for transparent encryption
VPS Virtual Private Server (opposed to shared hosting)

10 acronyms in this thread; the most compressed thread commented on today has 20 acronyms.

[Thread #292 for this comm, first seen 15th May 2026, 11:40] [FAQ] [Full list] [Contact] [Source code]

[-] gedfromgont@piefed.ca 1 points 1 day ago* (last edited 1 day ago)

I have a Pi 4B with 8 GB RAM and run Jellyfin plus some other stuff on it. Works great.

Had it first installed via repo but transcoding did not work at all. After switching to the Docker setup though, transcoding worked ok out of the box. Definitely takes a few seconds before the stream starts when having to transcode but no hiccups afterwards. Unless you jump around of course, and also I never had more than one stream trandcoding.

[-] SuspiciousCarrot78@aussie.zone 1 points 1 day ago* (last edited 1 day ago)

It's very ok, as long as you don't expect multiple 4K streams at it.

I ran JellyFin on a Pi 4 for about 3 or 4 yrs before it started acting up. So long as you don't transcode, it works wonderfully well. I had it serving upto 4-5 x 720p streams at same time. IIRC, it can just about do a single 4K, 60? Never tried - all my media is 1080p or less.

IIRC, mine is overclocked and undervolted using PiTools (and is in a Argon 40 case with a m.2). The Argon 40 case (I think) is causing it to short (something with the daughter-board? Dunno). Better options these days.

Paperless I don't use but I don't see why it shouldn't be possible.

Don't try Immich unless you like pain (or turn off the AI stuff)

[-] frosch@sh.itjust.works 1 points 1 day ago

Did you just turn off all transcoding or pre-transcoded to 720p mp4 (or similar)?

Or did you just rely on the native apps for direct play?

Paperless, I guess OCR is very taxing, but I mainly had RAM problems regarding this on my Pi 3, with only one doc at a time, it went slow but steady.

Immich maxed out my i5 (all cores on 100) for some minutes when running the ML on my imported test library of ca. 3k Files. So yeah, when choosing the Pi4, I'd go with ente for sure!

[-] SuspiciousCarrot78@aussie.zone 2 points 1 day ago* (last edited 1 day ago)

Yeah, transcoding entirely off - directly stream stored 720/1080p files (downloaded like that, although I did use handbrake on the pi once to transcode Space 1999 season 1. Took about 2 days I think).

Someone else was just talking about Wyse thin clients. I'm fairly sure that a $40 Wyse thin client out performs even the best Pi 4 (maybe 5 sometimes). If I can't find a way to fix mine, I may have to buy a few for uh...science. IIRC, they idle at about the same as the Pi

[-] frosch@sh.itjust.works 2 points 1 day ago

Yeah, I've read about using a Wyse 5070 for this a lot by now. Maybe I'll explicitly search for them instead. All I've seen around me are 4gb models, however. The Stack I'm currently running blocks 3.7gb in idle... So that's not enough

[-] Teppichbrand@feddit.org 3 points 2 days ago* (last edited 2 days ago)

I tinkered for years with numerous Raspberry Pies but got tired of it. Bought a second hand Dell ThinClient Wyse 5070 for like 70€, installed DietPi and its awesome.

[-] theorangeninja@sopuli.xyz 3 points 2 days ago

Consider papra as a more lightweight alternative for paperless-ngx. I have not used it yet unfortunately.

[-] frosch@sh.itjust.works 1 points 1 day ago

Gotta take a look at it and spin one container up. Looks promising, too!

As I'm not completely invested in paperless, I'll definitely try it out, thanks!

Haven't seen support for exports/backups at first glance, that is imo a must have.

this post was submitted on 14 May 2026
36 points (100.0% liked)

Selfhosted

59236 readers
724 users here now

A place to share alternatives to popular online services that can be self-hosted without giving up privacy or locking you into a service you don't control.

Rules:

  1. Be civil: we're here to support and learn from one another. Insults won't be tolerated. Flame wars are frowned upon.

  2. No spam posting.

  3. Posts have to be centered around self-hosting. There are other communities for discussing hardware or home computing. If it's not obvious why your post topic revolves around selfhosting, please include details to make it clear.

  4. Don't duplicate the full text of your blog or github here. Just post the link for folks to click.

  5. Submission headline should match the article title (don’t cherry-pick information from the title to fit your agenda).

  6. No trolling.

  7. No low-effort posts. This is subjective and will largely be determined by the community member reports.

Resources:

Any issues on the community? Report it using the report flag.

Questions? DM the mods!

founded 2 years ago
MODERATORS