ACE-Step-1.5
by ace-step
ACE-Step 1.5 is an open-source AI model that generates full songs, with vocals and lyrics, from a text description. It runs on your own computer and can also make covers, edit sections and separate existing audio into stems.
- Stars
- 14.1k
- Licence
- MIT
- Runs on
- A gpu machine, with a GPU
- Setup
- Moderate
- Language
- Python
- Age
- 418 days
What you can do with it
- Generate full songs of 10 seconds to 10 minutes from a short text description, with lyrics in 50+ languages
- Create covers, repaint sections, split tracks into stems or add backing music to a vocal recording
- Train a small LoRA on a few of your own songs so the output matches your style
Could replace
- Suno
- Udio
Our suggestion based on what it does, not a feature-by-feature comparison.
Make money with it
You could offer custom background tracks, jingles or personalised songs to local businesses and content creators.
An idea, not financial advice. Check the licence before charging for anything built on it.
Ways a business could use it
- Produce background music for videos, adverts, podcasts and social posts instead of paying for stock music subscriptions
- Make custom jingles, demo tracks or song drafts for clients as part of a creative or marketing service
Can you use it commercially?
Free to use, change and sell, including in paid products. Keep the copyright notice.
Not legal advice. Read the project's licence file before building a product on it.
Getting it running
Install uv, clone the repository, run "uv sync" and then "uv run acestep" to open a web interface, with models downloading on first run; Windows and Mac portable packages are also offered.
Follow the official instructions in the project README.
Things to check
- A CUDA GPU is recommended; the larger XL models need 12GB or more of VRAM, though smaller models can run on under 4GB
- Check the terms for the model weights and training data before using the music commercially
- Quality claims against commercial tools come from the project's own benchmarks
Who's talking about it
Checked 27 Oct 2026. Summary drafted by AI from the project's README.
