How to Setup BirdNET-Go on a Raspberry Pi
16 February 2026
BirdNET-Go is our preferred open-source software package to run real-time bird detection. It is a high-performance AI application that analyses live audio feeds 24/7 to identify, record, and visualize bird sounds.
This guide walks through the recommended hardware, storage, and microphone options, with step-by-step instructions to install a self-hosted BirdNET-Go instance on Raspberry Pi OS, so you can start monitoring quickly and reliably.
Getting the right equipment
To begin, you will need a machine to run the software and we recommend using a Raspberry Pi. Specifically, a Raspberry Pi 5 (8GB) is the safest choice; easily powerful enough to run the realtime inference, along with the web UI, audio handling etc. A Raspberry Pi 4/4B would also be fine (ideally 4GB or 8GB), but we have also run it on a Zero 2 W, so if you already have a less powerful machine like a 3B or Zero, then it’s still worth giving it a go, but in this case would certainly recommend using BirdNET-Go over BirdNET-Pi due to its efficiency.
In terms of picking a memory card, we’d recommend endurance cards. The Pi will be running 24/7 and doing a lot of writing to disk, as the audio is recorded. We also recommend going for a branded card, and ideally buy from a trusted vendor. There are lots of fakes on Amazon/Ebay, so if you buy there, buy direct from manufacturer stores. In terms of storage size, 128GB and above is recommended, just so you have more space to store audio records.
Our picks for microSD cards would be:
- SanDisk MAX Endurance 128GB+
- Samsung PRO Endurance 128GB+
- Kingston High Endurance 128GB+
Finally, you will need a microphone and we’ve found that the Samson Go Mic performed excellently for the price point. This isn’t an outdoor mic, so only good for home setups, where the mic is kept dry (under the eave of a roof or similar).
A great alternative DIY option is the Primo EM272J capsule paired with the Behringer UCA202 USB audio interface. The EM272 has ultra-low self-noise (14 dBA), making it excellent for passive bird monitoring. (Note that the capsule requires a small voltage divider circuit to power it correctly - this is well documented in the BirdNET-Pi community forums if you want to go down that route.)
Installing BirdNET-Go on the device
Next step is to set up the software required to run BirdNET. Firstly, we need to put an operating system on the memory card. The easiest way to do this is to use the Raspberry Pi Imager tool, which you can download for Windows and macOS here: https://www.raspberrypi.com/software/
Start the imager tool and follow the steps below:
- Select your device type from the list.
- Select an OS - we recommend Raspberry Pi OS Lite (64-bit), which is a more lightweight OS. This is not in the main options, so you need to scroll down and click ‘Raspberry Pi OS (other)’ to reveal more options.
- Select Raspberry Pi OS Lite (64-bit) from the list.
- Select your microSD card as the storage target.
- Before writing, the imager will offer to pre-configure settings. Click ‘Edit Settings’. Start by setting a hostname for your device - this is the name it will appear under on your network.
- Set your localisation options - timezone and keyboard layout - to match your region.
- Set a username and a strong password. You’ll use these to log in to the Pi.
- If you’ll be connecting the Pi over WiFi, enter your network name (SSID) and password here. If you’re using a wired ethernet connection, you can skip this.
- Enable SSH so you can connect to the Pi remotely from your computer without needing a monitor or keyboard attached.
- Once your settings are confirmed, click ‘Write’ to begin flashing the OS to the microSD card.
- Confirm you are happy to erase the card and proceed.
- Wait for the process to complete - the imager will verify the write automatically. Once done, you can safely remove the card.
Insert the microSD card into your Raspberry Pi, connect your microphone via USB, and power it on. Give it a minute to boot, then SSH into it from your computer using the hostname and credentials you set up in the imager.
From there you can follow the BirdNET-Go installation instructions to get the software running - it will automatically detect your USB audio interface and you’ll be up and listening in no time.
Note: the setup would be the same for other devices such as the Raspberry Pi 4, Zero 2 W etc - the only difference is that you select that device in the first step.