heartmula

Generate full MP3 songs from lyrics and tags using HeartMuLa models.

Updated Jun 17, 2026
One-click install
npx skills add https://github.com/anilcan-kara/nozich-agent --skill heartmula-anilcan-kara
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: heartmula
Source: https://github.com/anilcan-kara/nozich-agent/tree/main/skills/media/heartmula
Command: npx skills add https://github.com/anilcan-kara/nozich-agent --skill heartmula-anilcan-kara

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

HeartMuLa enables users to generate full songs from lyrics and tags using an open-source music engine, removing dependence on proprietary platforms.

Core Features & Use Cases

  • Music generation conditioned on lyrics and tags with HeartMuLa suite (HeartMuLaGen, HeartCodec, HeartTranscriptor, HeartCLAP).
  • Multilingual lyrics support and offline, local-generation workflows for open-source enthusiasts and indie composers.
  • Use Case: A creator writes a short lyric and tags and generates a complete MP3 track for production.

Quick Start

Install and run HeartMuLa locally to generate a song from a lyrics file and a tags file and save the output as an MP3.

Frequently Asked Questions about heartmula

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I generate full songs from lyrics and tags locally?

HeartMuLa generates full songs locally by using foundation models to process lyrics and tags files, outputting a complete MP3 track. This enables offline, open-source music creation without depending on proprietary platforms.

Can I use multilingual lyrics for AI music generation?

Multilingual lyrics are fully supported for AI music generation by conditioning the audio output on diverse language text inputs alongside tags. This satisfies the functional needs of multilingual lyric-conditioned song production.

Does open-source music generation work without an internet connection?

Open-source music generation works offline without an internet connection through local model loading and audio codec integration. This removes dependence on proprietary platforms and enables edge deployments for isolated workflows.

How do I minimize VRAM usage during GPU-accelerated audio generation?

To minimize VRAM usage during GPU-accelerated audio generation, the engine supports optional lazy loading. This defers loading model components until needed, significantly reducing memory footprint for local music production.

What do I need for offline lyric-conditioned song generation?

Offline lyric-conditioned song generation requires a lyrics file, a tags file, and a local GPU environment. The system integrates an audio codec and foundation models to synthesize and save the final MP3 track.