π ACE-Step v1.5 is a highly efficient open-source music foundation model designed to bring commercial-grade music generation to consumer hardware.
A local process extension by DrHepa. One Text to Music node uses the real ACE-Step 1.5 Turbo inference API to produce instrumental music or songs from a description and optional lyrics. This is music generation, not a speech TTS or voice-cloning adapter.
setup.py prepares an isolated runtime and downloads the pinned model files to Modly's persistent models directory. processor.py only reads local weights; it never downloads them. This follows the current process contract, not the separate weight-download UI used by model extensions.
Target: upstream Modly 0.4.2, audited at 1476fd0. See VALIDATION.md for the exact evidence and untested targets.
Worlds We Make β a two-minute epic synth-pop song about creating worlds with Modly, with original English lyrics and a requested female lead vocal.
https://github.com/user-attachments/assets/1a6ccf6c-75fa-4387-85ae-b88966342111
Read the full README on GitHub β