speaker-diarization-3.1
Model from pyannote that works out who speaks when in an audio recording.
AI model: the licence covers the model weights. Also check the model card (training data, limitations) and how you use the model under the EU AI Act.
Building on it: what may you do?
Short answers based on the licence, for the version in the catalog.
- B1 Change the code for internal use in your company Yes Yes. Keep the licence notices.
- B2 Build your own product or service on top and offer it to customers as SaaS Yes Yes. You don't have to share your code.
- B3 Make a closed-source derivative and sell or distribute it Yes Yes, even closed. The licence and copyright notices must ship with it.
- B4 Use the project's name and trademark No No, not without permission. The licence covers the code, not the name or logo.
Keep in mind
- There is a licence note for this product (e.g. parts under another licence or version-specific terms): read it and check the licence of the exact version you use.
This is guidance, not legal advice. Check with a lawyer before deciding.
The licence in brief
MIT License (MIT)
Permissive licence. You may use, modify and redistribute freely, including commercially. Keep the copyright and licence notice.
The model is gated on Hugging Face: you must sign in and accept terms before downloading. Read them; they may apply on top of the licence.
This is guidance, not legal advice (inte juridisk rådgivning). Check with a lawyer before deciding.
Facts
- Licence
- MIT
- Licence type
- Permissive
- Open core
- No
- GitHub stars
- Unknown
- Latest activity
- Deployment
- pyannote-audio
- Login/SSO
- Unknown
- API
- Unknown
- Multi-tenant
- Unknown
- EU vendor
- Unknown
- Source
- Hugging Face API, fetched 2026-10-03
Licence analysis per use case
Sign in to see what the licence means for eight use cases – from internal self-hosting to offering the tool as a service.
Sign in