HeartMuLa · Music

HeartMuLa

Commercial use permittedApache License 2.0

A 3B open-source music model, released with its own codec and transcriptor. The shipped catalog calls this its least verified entry, and that still holds for the numbers: every default below was copied from a sibling and not measured. The licence has been checked, and it is the most permissive of any music model in this catalog.

The licence

Summarized from the linked source documents, last checked 25 August 2026. Every claim links to its source; anything unconfirmed is marked unconfirmed.

Apache License 2.0

Commercial use permitted

Apache-2.0, per the upstream README and the license tag on the model repo; audio.cpp's conversion table agrees. Commercial use permitted with no threshold.

What about the audio you generate?

Apache-2.0 has no output clause. No separate output or acceptable-use document was found.

Correction to the shipped catalog: The shipped catalog says the licence is "Unknown" and the vendor "Unverified". Both have been improved on, with a caveat: the Apache-2.0 claim comes from the repository README and the Hugging Face licence tag, but the LICENSE file itself would not load, so the licence text has not been reviewed. No company behind the project could be identified; "HeartMuLa" is the only name it can be attributed to.

This licence has not been reviewed in full — the label above is the one the model's publisher applies, repeated. Treat it as a pointer, not an answer.

Specs

Transcribed from the shipped catalog, audiocpp_models_catalog.json — that file is authoritative if a number here disagrees with it.

Vendor
HeartMuLa
Catalog id
audiocpp-heartmula
audio.cpp loader family
heartmula
Can do
Generation
Weights file
HeartMuLa-GGUF/heartmula-q8_0.gguf
Account needed to download
No
Max clip length
60s
Sample rate
44.1kHz stereo
Approx. VRAM
4.0 GB
In the picker by default
No — opt in from Settings
Run end-to-end here
No — catalogued, not tested here

Getting it

There is no per-model install. One native binary runs all of these, so getting a model is a download of its weights and nothing else — no virtualenv, no torch pin, no CUDA wheel to match against your driver.

From the CLI
soundbox model install audiocpp
soundbox pull audiocpp-heartmula
  • The binary is fetched from audio.cpp's GitHub releases the first time you ask for it and cached under your own data directory. Soundbox ships no native binaries and vendors none.
  • Weights land in <models dir>/audiocpp/audiocpp-heartmula/ rather than in the shared Hugging Face cache, so a download stays scoped to the model you asked for.
  • This one is opt-in: switch it on in the Models page's “Enable more models” section before it appears in the picker.

Worth knowing

  • Every numeric default in the specs table is inherited, not measured. Check them against the model card before real use.
  • Opt-in: switch it on in Settings before it appears in the picker.

Sources

Primary sources: the model's own repository, the project it came from, and the conversion audio.cpp actually downloads. Never invented — an entry with nothing findable says so instead.

Licence details on these pages are summaries of the linked source documents, last checked 25 August 2026. They are not legal advice, and Soundbox is not affiliated with any model's authors — weights download from Hugging Face under the user's own account and terms. Company and model names identify whose work it is, and belong to their owners. The linked licence text is authoritative; commercial use should be checked against it, per model, before anything ships.