Featured by Willet
What it is
A Gradio web interface that chains together YouTube downloading, Demucs vocal separation, Whisper-based transcription, machine translation into 100+ languages, and speech synthesis with Edge-TTS, kokoro, F5-TTS, E2-TTS and CosyVoice. It can clone a voice from a short sample and produce dubbed audio and subtitles. Podcasters, video creators and translators run it on their own machine, mainly Windows with an NVIDIA GPU, to localize content without a paid voice service. Active development is currently paused.