Cuelora Pro Open-Source and Third-Party Model Notices Last updated: October 17, 2026 This file summarizes third-party open-source software and model assets used by the Cuelora Pro desktop application and the Cuelora patron applications. These notices do not change the Cuelora Pro Software License Agreement. Each third-party component remains licensed under its own license. ======================================================================== Part 1 - Base Cuelora Pro Desktop Application Open-Source Components ======================================================================== The Cuelora Pro installer uses Python packages and tools including: - Flask - BSD-3-Clause - Flask-SocketIO - MIT - python-socketio / python-engineio - MIT - Socket.IO JavaScript client - MIT - NoSleep.js 0.12.0 - MIT, Copyright (c) Rich Tibbett. Unmodified local distribution and license: static/vendor/nosleep/NoSleep.min.js and LICENSE. Includes embedded video media for the patron keep-awake fallback. - qrcodejs - MIT - simple-websocket - MIT - pdfplumber - MIT - pdfminer.six - MIT - python-docx - MIT - pyserial - BSD - cryptography - Apache-2.0 / BSD - SpeechRecognition - BSD - sentencepiece - Apache-2.0 (tokenisation for offline OPUS-MT translation) - soundfile / python-soundfile - BSD-3-Clause - libsndfile (audio file library used by soundfile) - LGPL-2.1-or-later, bundled as a dynamically linked shared library inside the soundfile binary wheel - LAME (MP3 encoder, used to store generated program audio as MP3) - LGPL-2.0-or-later (GNU Library General Public License), compiled into the libsndfile shared library - mpg123 / libmpg123 (MPEG audio decoder) - LGPL-2.1, compiled into the libsndfile shared library - libFLAC, libogg, libvorbis, and Opus (Xiph.Org audio codecs) - BSD-3-Clause, compiled into the libsndfile shared library - Apple Lossless (ALAC) codec code - Apache-2.0, compiled into the libsndfile shared library - GSM 06.10 and G.72x ADPCM codec code - under their own permissive notices, compiled into the libsndfile shared library - pyttsx3 - open-source Python package for platform speech APIs - faster-whisper - MIT - Silero VAD asset bundled by faster-whisper for local voice activity detection - MIT - CTranslate2 - MIT (runs faster-whisper and the offline OPUS-MT translation models) - onnxruntime - MIT - NumPy - BSD-style license - Pillow - HPND - pystray - LGPL-3.0 - PyInstaller - GPL-2.0-or-later with PyInstaller bootloader exception - PyAV - BSD-3-Clause (bundled as a dependency of faster-whisper) - FFmpeg libraries (libavcodec, libavformat, libavfilter, libavdevice, libavutil, libswresample) - LGPL-2.1-or-later, bundled as dynamically linked shared libraries inside the PyAV binary wheel - dav1d (AV1 decoder, bundled with PyAV) - BSD-2-Clause - SVT-AV1 (AV1 encoder, bundled with PyAV) - BSD-3-Clause-Clear ======================================================================== Part 2 - Cuelora Patron Applications (Android and iOS) ======================================================================== - ZXing Core (`com.google.zxing:core`) - Apache-2.0 (used in the Cuelora Android patron application for local QR code decoding; camera frames are analyzed in memory on the device and never stored or transmitted) - AndroidX Core, AppCompat, Activity, WebKit, CameraX (`camera-core`, `camera-camera2`, `camera-lifecycle`, `camera-view`) and Google Material Components for Android - Apache-2.0 ======================================================================== Part 3 - Third-Party Model Notices (OPUS-MT and Silero VAD) ======================================================================== Offline translation models (OPUS-MT) ------------------------------------ Cuelora Pro 5.2.3 translates scripts on the licensed computer with OPUS-MT neural translation models published by the University of Helsinki (Tatoeba Challenge releases, https://github.com/Helsinki-NLP/Tatoeba-Challenge), run locally with CTranslate2 (MIT) and SentencePiece (Apache-2.0). Supported offline language pairs from English: Spanish (es), French (fr), Italian (it), Portuguese (pt), Arabic (ar), and Korean (ko). OPUS-MT models: Copyright University of Helsinki, Language Technology Research Group. Licensed under the Creative Commons Attribution 4.0 International License (CC BY 4.0, https://creativecommons.org/licenses/by/4.0/). Attribution: - Tiedemann, J. and Thottingal, S. (2020). "OPUS-MT - Building open translation services for the World." Proceedings of the 22nd Annual Conference of the European Association for Machine Translation (EAMT). - Tiedemann, J. (2020). "The Tatoeba Translation Challenge - Realistic Data Sets for Low Resource and Multilingual Machine Translation." Proceedings of the Fifth Conference on Machine Translation (WMT). No translation model file is included in the installer. An administrator downloads each language pair once (~1 GB per language) from the Helsinki-NLP release server (object.pouta.csc.fi) under Settings > AI Engines > Translation, between shows; the model is converted to a compact CTranslate2 form and kept locally under the data folder (models\opus-mt). The original model card is kept beside it as MODEL_README.md. Script text processed by these models never leaves the computer. No NLLB or other non-commercial model is used. The unofficial Google Translate endpoint used by earlier versions has been removed. Optional bring-your-own-key translation (DeepL) sends dialogue lines (and uploaded show program text when its translate option is selected) to that provider under the customer's own account and the provider's terms, and only when the customer configures a paid API key. Base installer model asset (Silero VAD) --------------------------------------- The Cuelora Pro installer includes `silero_vad_v6.onnx` only as a voice activity detection asset through faster-whisper. It is used for local live speech detection, not for text-to-speech generation, voice cloning, or voice creation. Silero VAD is published under the MIT License. ElevenLabs BYOK note -------------------- Cuelora Pro does not bundle ElevenLabs model files, ElevenLabs voice data, or any ElevenLabs API key. ElevenLabs support is an optional bring-your-own-key cloud text-to-speech integration for rehearsal and prebuilt script audio and for read-aloud audio of an uploaded show program, only when configured by the customer. ======================================================================== Part 4 - Standard Open-Source License Notices ======================================================================== MIT License notice ------------------ Permission is hereby granted, free of charge, to any person obtaining a copy of software and associated documentation files licensed under the MIT License, to deal in the software without restriction, including without limitation the rights to use, copy, modify, merge, publish, distribute, sublicense, and/or sell copies of the software, subject to the license's copyright notice and permission notice being included in all copies or substantial portions of the software. Apache License 2.0 notice ------------------------- Apache-2.0 licensed components (including SentencePiece and ZXing Core) are provided under the Apache License, Version 2.0. You may obtain a copy of the license at: https://www.apache.org/licenses/LICENSE-2.0 Creative Commons Attribution 4.0 International (CC BY 4.0) notice ----------------------------------------------------------------- OPUS-MT neural translation models are licensed under the Creative Commons Attribution 4.0 International License (CC BY 4.0). You may obtain a copy of the license at: https://creativecommons.org/licenses/by/4.0/ BSD-style notices ----------------- BSD-licensed components are provided under their respective BSD license terms, including the requirement to retain copyright, license, and disclaimer notices. LGPL notices (FFmpeg libraries, pystray, libsndfile, LAME, mpg123) ------------------------------------------------------------------ The FFmpeg libraries listed above are licensed under the GNU Lesser General Public License version 2.1 or later (LGPL-2.1-or-later). pystray is licensed under the GNU Lesser General Public License version 3.0 (LGPL-3.0). libsndfile is licensed under LGPL-2.1-or-later, LAME under the GNU Library General Public License version 2 or later (LGPL-2.0-or-later), and mpg123 under the GNU Lesser General Public License version 2.1 (LGPL-2.1). Copies of the LGPL are available at: https://www.gnu.org/licenses/old-licenses/lgpl-2.0.html https://www.gnu.org/licenses/lgpl-2.1.html https://www.gnu.org/licenses/lgpl-3.0.html These components are distributed as separate, dynamically linked files inside the Cuelora Pro install directory (the FFmpeg libraries as shared library files under the PyAV component, pystray as separate Python module files, libsndfile as a shared library file under the soundfile component, with LAME and mpg123 compiled into that libsndfile file). Users may replace them with modified versions of the same libraries, as permitted by the LGPL. FFmpeg is a trademark of Fabrice Bellard, originator of the FFmpeg project; source code for FFmpeg is available at https://ffmpeg.org. Source code for pystray is available at https://github.com/moses-palmer/pystray. Source code for libsndfile is available at https://github.com/libsndfile/libsndfile, for LAME at https://lame.sourceforge.io, and for mpg123 at https://www.mpg123.de. To obtain the complete corresponding source code for the LGPL-licensed library versions distributed with Cuelora Pro, contact support@cuelora.com; Cuelora will provide it at no more than the cost of distribution. This offer is valid for three years from the date of distribution of the Software. PyInstaller bootloader exception -------------------------------- PyInstaller is licensed under GPL-2.0-or-later with an exception that permits distribution of the PyInstaller bootloader as part of non-GPL applications. That exception applies only to PyInstaller and does not change the license of Cuelora Pro proprietary code. Requests -------- For a copy of notices, source-offer information where applicable, or questions about third-party components, contact support@cuelora.com.