debpalash/OmniVoice-Studio — reverse-engineered prompt

Build me a desktop app that works like a local ElevenLabs alternative. I want people to run it on their own computer with no API keys, no accounts, and no cloud upload.

The app should let a user drop in a short voice clip and clone that voice, type text and generate speech, and design a new voice by choosing things like age, gender, accent, pitch, speed, emotion, and language. It should also handle video dubbing, where someone can paste a YouTube link or upload a video, transcribe it, translate it, create a new voice track, and export an MP4.

Please include a simple dictation widget that can be opened from any app with a keyboard shortcut, records speech, transcribes it, pastes the text, then gets out of the way. Add a searchable library for voices, projects, exports, and history, plus settings for installing models, checking device support, and viewing logs.

Make the interface polished and beginner friendly. Look up current docs online if you need to.