I’m astounded that I cannot find a single thing on this?

From my understanding, the Raspberry Pi 5 has its GPIO mapped differently or on a different chip?

The HACS rpi gpio integration does not seem to work and list the new entities with RPi5.

How can I utilise the RPi5 GPIO pins and set it up for simple on/off function?

I’m using HASSOS

Any help would be amazing.

I’m unsure of how to implement that with HASSOS on the Pi5?
I tried the code straight into my config.yaml but I’m sure there’s more needed?

I’m an absolute noob.

Unfortunately i do not use HASSOS nor an rpi as I am using docker. Also the link I sent may be for a remote RPI. So may need help from others on the forum. What specifically are you attempting?

I’m just upgrading from Pi4 to Pi5 and was previously using the GPIO to have hardwired control of relays to switch things on and off in a van home.

Upgrade was easy but haven’t for the life of me been able to get the old HACS RPi GPIo integration to work and show entities

I’ve done a clean install and still nothing.
All other functions and HACS integrations are fine.
It’s just the GPIO.

The integration hasn’t been updated since before the Pi5 so I believe it’s a compatibility issue. Someone also posted on the GitHub an issue request of the same problem.

So I’m looking for another integration or code to work around

If all else fails you can hack something together by directly interacting with the pigpiod daemon. I’m assuming it would still be on a Pi 5.

Also check if you can find something on sparkfun.com.

What about the MQTT IO add-on from the add-on store?

Expose GPIO modules and digital sensors via MQTT for remote control and monitoring.

Any solution? I am stuck on the same problem.

I haven’t found one yet. Pigpio I believe has the same issue as the RPi GPIO integration. The Pi5 seems to have setup its GPIO differently so none of the current methods seem to work.

I have tried MQTT yet.

If that’s the case then your luck is likely out until that is resolved, unfortunately.



Can’t believe no one has worked it out.
GPIO on the Pi is clearly not as popular as I thought

It’s not trivial work. Did you check the links I shared? The hardware architecture changed with the Pi 5. There is an underlying new library, but all implementations using the older libs need to change. This takes time and effort — probably people’s free time. Also, the Pi is falling out of popular demand, mostly due to price and alternatives.

Yeh I did. I understand it’s changed, what a pain in the neck. I think I might move to a mini PC and use the old Pi4 as remote gpio

My long-term plan is to migrate over to an ESP8266 running ESPHome. It’s already set up and monitoring a few temperature probes around my boiler, so it’s just a matter or moving all the wires over from the RPi. Not convenient, but it’s pretty obvious that GPIO on the RPi isn’t the way to go. It’s too bad they’ve priced the RPi out of the market.

I think the increasing lack of support is also what kills it. Personally I don’t mind the money, it’s still cheaper than tons of options, especially the proprietary eco systems and hubs. And it just works.
But the fact it’s now been out 6 months and no one has a solution for GPIO is disappointing and what kills the RPi for me.

When the Pi4 came out with no support for SSD boot, people had to wait a year before this was implemented in beta (and a while longer until it was stable).

That’s the downside of buying the newest products. You can’t be “disappointed” because something you want isn’t implemented within the first 6 months. It’s your responsibility to do your own research before buying a product, rather than complain that it kills it for you.


In the open-source world, there’s no support because nobody wants to volunteer their time to support it. I don’t blame the RPi for that!

It’s frustrating to me to have those perfectly functional GPIO pins sitting there, unused, right next to the bank of relays I want to read from. But I accept the fact that, for whatever reason, very few people want to do what I’m trying to do. Why would anyone put all that (unpaid) work into supporting something hardly anyone uses? If I cared enough, I could do it myself.

Or, give in to reality and find a different, albeit less efficient, method. The fact is, there’s no-one to blame, or to complain to.

1 Like

Fair point! Agree with that statement. I’ll work around it.