BirdNET-GO CLI Usage

[hpk60]

On the first page, there are instructions about the usage of the CLI, however, I do not have the birdnet command installed.
Do I need to install additional code to use the birdnet commands?

`BirdNET-Go CLI

Usage:
birdnet [command]

Available Commands:
authors Print the list of authors
benchmark Run performance benchmark
directory Analyze all *.wav files in a directory
file Analyze an audio file
help Help about any command
license Print the license of Go-BirdNET
rangefilter Manage the range filter database
realtime Analyze audio in realtime mode
support Generate a support bundle`

[tphakala]

Oh that is an error in documentation, command should be birdnet-go, not birdnet. Thanks for pointing that out!

Also it is important to note that cli tool is not available if you have done container based install using install.sh, to have access to cli you need to install standalone binary for your host os, they are available on each release https://github.com/tphakala/birdnet-go/releases/tag/nightly-20250731

If you are using Linux you need to copy tensorflow lite library file libtensorflowlite_c.so into library path like /usr/local/lib and run sudo ldconfig, after that it should work.

cp libtensorflowlite_c.so /usr/local/lib
sudo ldconfig

[hpk60]

Thank you @tphakala

If using the CLI, will detections with "birdnet-go directory" be added to the database like the ones from the microphone?
Are they visible also on the birdweather homepage like the ones taken with the microfone?

[tphakala]

No, it will only report data in text format.

It is basically a replacement for BirdNET Analyzer command line tool (https://github.com/birdnet-team/BirdNET-Analyzer), meant to analyse hours long field recordings.

[hpk60]

okay, I see ... I was experimenting with the BirdNET-Analyzer with 2GB files from a Zoom microphone with good results.
Here BirdNET-Analizer runs on the PC, not on the raspberry pi. Good to know, that there are alternatives :-)