Every Lector

Open source

Open-source credits

Every Lector stands on open-source work. Here is what ships inside it and under which licence.

ComponentUsed forLicence
faster-whisper and CTranslate2Speech recognitionMIT
Whisper model weightsSpeech recognition modelsMIT
Chatterbox MultilingualVoices in 20 languages and voice cloning, with an inaudible watermarkMIT
KokoroFast voices for common languagesApache-2.0
Qwen3Translation and subtitle shorteningApache-2.0
llama.cppRuns the translation model on your PCMIT
pyannote community-1Telling speakers apartCode MIT, weights CC-BY-4.0 (attribution)
FFmpeg (LGPL build)Reading and writing video and audioLGPL
FlutterThe app interfaceBSD
media_kit and libmpvVideo playbackLGPL, dynamically linked

Full licence texts ship with the installer. Product names belong to their owners.

Film still

The hero preview uses a frame from Sintel (c) Blender Foundation, durian.blender.org, licensed CC BY 3.0. The subtitles shown over it are an illustration, not lines from the film.