ACE-Step ships open-source music generation foundation model
The Apache-2.0 text-to-audio model turns a text prompt and lyrics into complete songs in seconds — and anyone can download, fine-tune, and build on it.
By Ryan Merket · Published
Primary source: ACE-Step
ACE-Step, an open-source foundation model for music generation, is now available at https://ace-step.club/ . The model turns a text prompt plus optional lyrics into complete songs of up to four minutes in seconds, and runs locally under the Apache-2.0 license.
Unlike prompt-to-audio demos, ACE-Step is built as a foundation model: fast, steerable, and open end to end. It ships checkpoints, inference code and a LoRA training pipeline, so musicians and developers can fine-tune it and ship products on top. Capabilities include text-to-music, editing, remixing, voice cloning and stem splitting.…
When you purchase through links in our articles, we may earn a small commission. This doesn’t affect our editorial independence.