Skip to content

DiariZen

Speaker diarisation powered by Pyannote 3.1

Visit official site

What DiariZen is

The toolkit is described as a speaker diarisation package driven by AudioZen and Pyannote 3.1.

What you can do with it

  • Separate speakers in audio
  • Use public datasets for evaluation
  • Customise the open-source code

Application scenarios

  • Study meeting speech patterns
  • Process surveillance recordings
  • Embed diarisation in an app

Pros & cons

✓ What we like

  • Commercial use is stated as allowed
  • WavLM Base+ and ResNet34-LM are supported

! What to watch out for

  • Setup needs a virtual Python environment
  • Datasets and pre-trained models must be downloaded

FAQ

What license is used?

The material states the MIT license on GitHub.

What is it built on?

The material states AudioZen and Pyannote 3.1.

Last reviewed: 2026-11-15

More AI audio tools tools

View all →

How we review