MuMa (Music Maker)

About MuMa

MuMa (Music Maker) is an open-source AI music generation system combining:

• ACE-Step: Diffusion-based music synthesis model
• LLM-powered lyrics: GPT-2/Phi-3 based lyric generation
• Whisper: Voice input and reference processing
• Voice Cloning: Optional voice reference for personalized vocals

Features

• ¼ Text-to-music generation
• ¤ Vocal and instrumental support
• 🗣️ Voice recording and cloning
• 📊 Multi-format export (WAV, FLAC, OGG, MP3)
• Melody composition aligned with lyrics

License

Apache 2.0 License

Part of the LEMM (Let Everyone Make Music) Project
Making high-quality AI-generated music available to everyone for free.

© 2025 Gamahea (Creator) | Claude Sonnet 4.5 (Lead Developer)

System Requirements

• Python 3.9+
• 8GB+ RAM (16GB recommended)
• GPU optional (CPU mode supported)

Audio Format Support

• Input: WAV, MP3, FLAC, OGG
• Output: WAV, FLAC, OGG (MP3 requires additional codec)