Open source, clearly credited
The code, model files, hosted service, and your recording have different rights and responsibilities.
WhisperX code
The upstream m-bain/whisperX repository publishes its code under BSD-2-Clause. Its copyright, license conditions, and disclaimer must be preserved when the code is redistributed. Read the upstream LICENSE for the exact terms. Paying for this website purchases hosted processing and related service features; it does not purchase ownership of the open source project.
Model files and dependencies
A code license does not automatically cover every downloaded model, dataset, or dependency. Speech recognition, alignment, and speaker models carry their own notices and access conditions. Speaker diarization is powered by pyannoteAI’s speaker-diarization-community-1 model, licensed under Creative Commons Attribution 4.0 (CC BY 4.0). This attribution concerns the speaker model; it does not assign that license to the other models or the entire service.
pyannoteAI speaker-diarization-community-1 model card ↗Your content
You retain the rights you already have in your uploaded recording. You grant the service the limited permission needed to store, process, and return the content. You are responsible for permission to upload and for any required consent from recorded participants. A transcript does not grant new rights in the underlying recording.
Independent identity
The WhisperX name describes the technology used by this independent service. No affiliation or endorsement is implied. Our website design and interface are separate from upstream branding. The authors of WhisperX, OpenAI, and pyannote are not responsible for this service.
Notices and questions
Upstream sources: https://github.com/m-bain/whisperX and its LICENSE; speaker model information: https://huggingface.co/pyannote/speaker-diarization-community-1. Existing commercial template licenses and third party notices are preserved in the software repository. Contact support@whisperx.app with licensing or rights concerns.
Speaker model attribution: pyannoteAI · CC BY 4.0 license terms.
Read the upstream WhisperX project · Contact support