v

voice-pro

Developed by abus-aikorea

Voice-Pro is a state-of-the-art, AI-powered web application designed for comprehensive multimedia processing, serving as a robust open-source alternative to ElevenLabs. Accessible via a Gradio interface, it seamlessly integrates YouTube video downloading, voice separation, and top-tier speech recognition using the Whisper model family. Its standout feature is zero-shot voice cloning powered by F5-TTS, E2-TTS, and CosyVoice, combined with instant translation and multilingual TTS supporting over 100 languages. Recently made completely free and open-source, Voice-Pro empowers creators, researchers, and developers to process audio efficiently via local desktop deployment, especially optimized for Windows with NVIDIA GPUs.

  • High-accuracy speech recognition via Whisper and WhisperX
  • Zero-shot voice cloning powered by F5-TTS and CosyVoice
  • Instant translation and multilingual TTS for 100+ languages
  • Automated YouTube processing and vocal separation workflow
desktopweb