HushScribe - Third-Party Notices ================================ HushScribe is built with the following open-source software. We are grateful to their authors. Speech recognition - Whisper models (tiny, base, large-v3-turbo) - OpenAI - MIT License (converted to 8-bit CTranslate2 format) https://github.com/openai/whisper - faster-whisper - SYSTRAN - MIT License https://github.com/SYSTRAN/faster-whisper - CTranslate2 - OpenNMT - MIT License https://github.com/OpenNMT/CTranslate2 - Silero VAD (voice activity detection, bundled with faster-whisper) - MIT License https://github.com/snakers4/silero-vad - ONNX Runtime - Microsoft - MIT License https://github.com/microsoft/onnxruntime - Hugging Face Tokenizers - Apache License 2.0 https://github.com/huggingface/tokenizers Live audio capture - SoundCard - Bastian Bechtold - BSD 3-Clause License https://github.com/bastibe/SoundCard - cffi - MIT License Audio / video decoding - PyAV - BSD 3-Clause License https://github.com/PyAV-Org/PyAV - FFmpeg libraries (dynamically linked, unmodified, shipped with PyAV) - GNU Lesser General Public License v2.1 or later This software uses libraries from the FFmpeg project under the LGPLv2.1. The FFmpeg source code is available at https://ffmpeg.org/download.html and the exact build used by PyAV at https://github.com/PyAV-Org/pyav-ffmpeg The FFmpeg DLLs are separate files in the application folder and may be replaced. User interface and runtime - pywebview - BSD 3-Clause License https://github.com/r0x0r/pywebview - pythonnet - MIT License https://github.com/pythonnet/pythonnet - Microsoft Edge WebView2 Runtime (part of Windows; not redistributed) - Python 3.12 - Python Software Foundation License - NumPy - BSD 3-Clause License - Pillow - MIT-CMU License - PyInstaller bootloader - GPLv2 with bootloader exception (permits commercial distribution) Full license texts are available at the project URLs above.