ACE-Step 1.5: Pushing the Boundaries of Open-Source Music Generation
Paper โข 2602.00744 โข Published โข 12
Model
This model is hosted for Demodokos Foundry but it can be used for other purposes, enjoy a stable download location and custom quantizations not available elsewhere.
ACE-Step 1.5 XL โ Base (4B DiT)
Project | Hugging Face | ModelScope | Space Demo | Discord | Tech Report
Model Details
This is the XL (4B) Base variant of ACE-Step 1.5 โ a larger DiT decoder with ~4B parameters for higher audio quality. It is the foundation model supporting all tasks: text-to-music, cover, repaint, extract, lego, and complete.
XL Architecture
| Parameter | Value |
|---|---|
| DiT Decoder hidden_size | 2560 |
| DiT Decoder layers | 32 |
| DiT Decoder attention heads | 32 |
| Encoder hidden_size | 2048 |
| Encoder layers | 8 |
| Total params | ~4B |
| Weights size (bf16) | ~18.8 GB |
| Inference steps | 50 (with CFG) |
GPU Requirements
| VRAM | Support |
|---|---|
| โฅ12 GB | With CPU offload + INT8 quantization |
| โฅ16 GB | With CPU offload |
| โฅ20 GB | Without offload |
| โฅ24 GB | Full quality (XL + 4B LM) |
All LM models (0.6B / 1.7B / 4B) are fully compatible with XL.
Key Features
- ๐ฐ Commercial-Ready: Trained on legally compliant datasets. Generated music can be used for commercial purposes.
- ๐ Safe Training Data: Licensed music, royalty-free/public domain, and synthetic (MIDI-to-Audio) data.
- ๐ฏ Full Task Support: Text2Music, Cover, Repaint, Extract, Lego, Complete.
- ๐ฎ Higher Quality: 4B parameters provide richer audio quality compared to the 2B variants.
Quick Start
# Install ACE-Step
git clone https://github.com/ace-step/ACE-Step-1.5.git
cd ACE-Step-1.5
pip install -e .
# Download this model
huggingface-cli download ACE-Step/acestep-v15-xl-base --local-dir ./checkpoints/acestep-v15-xl-base
# Run with Gradio UI
python acestep --config-path acestep-v15-xl-base
Model Zoo
XL (4B) DiT Models
| DiT Model | CFG | Steps | Quality | Diversity | Tasks | Hugging Face | ModelScope |
|---|---|---|---|---|---|---|---|
acestep-v15-xl-base |
โ | 50 | High | High | All (extract, lego, complete) | This repo | Link |
acestep-v15-xl-sft |
โ | 50 | Very High | Medium | Standard | Link | Link |
acestep-v15-xl-turbo |
โ | 8 | Very High | Medium | Standard | Link | Link |
2B DiT Models
| DiT Model | CFG | Steps | Hugging Face | ModelScope |
|---|---|---|---|---|
acestep-v15-turbo (default) |
โ | 8 | Link | Link |
acestep-v15-sft |
โ | 50 | Link | Link |
acestep-v15-base |
โ | 50 | Link | Link |
LM Models (all compatible with XL)
| LM Model | Params | Audio Understanding | Composition | Hugging Face | ModelScope |
|---|---|---|---|---|---|
acestep-5Hz-lm-0.6B |
0.6B | Medium | Medium | Link | Link |
acestep-5Hz-lm-1.7B |
1.7B | Medium | Medium | Included in main | Included in main |
acestep-5Hz-lm-4B |
4B | Strong | Strong | Link | Link |
Acknowledgements
This project is co-led by ACE Studio and StepFun.
Citation
@misc{gong2026acestep,
title={ACE-Step 1.5: Pushing the Boundaries of Open-Source Music Generation},
author={Junmin Gong, Yulin Song, Wenxiao Zhao, Sen Wang, Shengyuan Xu, Jing Guo},
howpublished={\url{https://github.com/ace-step/ACE-Step-1.5}},
year={2026},
note={GitHub repository}
}
- Downloads last month
- 58