Open-source foundation model for music generation with SOTA performance.
ACE-Step is a cutting-edge, open-source foundation model for music generation. It turns text prompts and lyrics into complete songs with vocals, delivering state-of-the-art quality through a fast, efficient architecture. Musicians, producers, content creators and developers can generate, remix and build on top of it with fully open weights and code. Main features: text-to-music generation with vocals and lyrics, open weights for self-hosting and fine-tuning, multi-genre and multi-language output, and a fast architecture designed for real-world production use. Ideal customer: independent musicians, studios, game developers, app builders and anyone embedding music generation into their products.