Instructions to use ACE-Step/Ace-Step1.5 with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Transformers
How to use ACE-Step/Ace-Step1.5 with Transformers:
# Use a pipeline as a high-level helper from transformers import pipeline pipe = pipeline("text-to-audio", model="ACE-Step/Ace-Step1.5", trust_remote_code=True)# Load model directly from transformers import AutoModel model = AutoModel.from_pretrained("ACE-Step/Ace-Step1.5", trust_remote_code=True, dtype="auto") - Notebooks
- Google Colab
- Kaggle
IMHO This is SUNO 3.5 or 4 but OPENSOURCE
I think this version can compare to the old version of SUNO. How can a small model be this good!?
I agree, most songs I generated sound very similar ... like you are using some MIDI with a good SoundFont plus one female and one male voice.
Anyhow I am very impressed what the devs packed into this small package.
Holy cow, it even works fine in cpu only mode.
С музыкой намного проще чем с изображение или видео,
If anyone wants to try ACE-Step without setting up a local environment, check out Boppy — it's a free web app powered by ACE-Step 1.5. You describe a song in plain text, an LLM writes the lyrics, and ACE-Step generates the full audio track. Any genre, up to 5 minutes, no signup needed. Great way to experience what this model can do without any installation.