HeartMuLa at a glance
Checked 16 Oct 2026HeartMuLa is a family of open music models released in January 2026: a 3B music language model that sings lyrics in many languages to style tags, the HeartCodec audio codec, and a lyrics transcriber. Code and all weights moved to Apache-2.0 in January 2026. It runs from a Python script, writes up to 4 minutes by default, and runs at about real-time speed. There is no UI, reference-audio input or section editing yet; covers live in a separate MuLaCover repo. Hardware and OS support are not documented, and the 7B model is not released.
RepoQuarry. “HeartMuLa.” Checked 16 Oct 2026. https://repoquarry.dev/projects/heartmulaHow much of Suno does it replace?
1 of 5 jobs. Song generation from lyrics and tags only, from a script; no UI, covers, edits or stems in this repo.
For instrumental tracks (Instrumental mode), covers and restyles of an uploaded song (Cover), extending or regenerating part of a song (Extend, splitting a song into vocal and instrument stems (Get Stems), see other Suno alternatives. You may also need to build an interface, an API server and hosting yourself.
Facts, with sources
Every answer links to where we found it. We read the documentation; we did not install it.
Clone this repo and install locally.github.com · heartlib#%EF%B8%8F-local-deployment ↗ 1d ago
This repository is licensed under the Apache 2.0 License.github.com · heartlib#%EF%B8%8F-license ↗ 1d ago
We update the license of this repo and all related model weights to Apache 2.0.github.com · heartlib#-news ↗ 1d ago
We recommend using `python=3.10` for local deployment.github.com · heartlib#-local-deployment ↗ today
Download our pretrained checkpoints from huggingface or modelscope using the following commandgithub.com · heartlib#%EF%B8%8F-environment-setup ↗ 1d ago