Speech recognition
Speech recognition in TutDub is powered by NVIDIA Parakeet TDT 0.6b v3, used under the Creative Commons Attribution 4.0 International License (CC BY 4.0).
Model: nvidia/parakeet-tdt-0.6b-v3. NVIDIA and the NVIDIA logo are trademarks of NVIDIA Corporation.
Voice cloning
Translated speech is re-synthesized in the original speaker's voice using VoxCPM, © OpenBMB, licensed under the Apache License 2.0.
Repository: github.com/OpenBMB/VoxCPM.
Other open-source software
TutDub also uses other open-source components, including FFmpeg (licensed under the GNU GPL; source code is available at ffmpeg.org/download.html).
Speech recognition outside Parakeet's language coverage runs on faster-whisper, © SYSTRAN, licensed under the MIT License, with OpenAI Whisper model weights, © OpenAI, MIT License.
Repository: github.com/SYSTRAN/faster-whisper.
A complete list of open-source components and their licenses is available upon request — please contact us.