I Built a Local AI Music Studio

Description and use

Miso is a local AI music studio that integrates generation, repainting, stem separation, voice conversion, and MIDI transcription into a single, traceable workflow.

Problem

Users want an integrated local environment to generate, remix, and edit music with AI models without relying on command-line tools or external services.

Outcome

Miso streamlines local AI music production, providing a cohesive workflow from prompt to final mix, with full traceability and offline operation.

Workflow

• Install Miso and audio.cpp via Docker Compose
• Create a new project and import a prompt or audio file
• Use the guided prompt builder to generate a full song with ACE‑Step
• Select a waveform region and repaint that section
• Split the track into stems and adjust levels or swap vocals
• Transcribe a vocal stem to MIDI, convert to ABC score, and feed to YuE2 for a new arrangement
• Export the final mix as a 48 kHz stereo file
• Save the project to preserve all takes and lineage

Evidence

Miso is a self-hosted local music generation and remix studio built on top of audio.cpp.

Source

The whole thing runs as two Docker containers.

Source