Descriptions:
Ace-Step 1.5 is a free, open-source music generation model positioning itself as a local alternative to subscription services like Suno and Udio. In this tutorial from the Nerdy Rodent channel, the host walks through two complete ComfyUI workflow setups for running the model on a home PC, generating full songs with custom lyrics and style tags in roughly eight seconds. For users without a powerful GPU, a Hugging Face Space is also available for free cloud-based inference.
The video covers a basic checkpoint workflow (requiring a single 10 GB model download) and a more advanced “rodent method” using KJ nodes to keep the ComfyUI graph clean and modular. Configurable parameters include song duration, BPM, key scale, seed, and language. A standout section demonstrates an AI-enhanced version that integrates Ollama — running Gemma 34B locally — to automatically generate song lyrics from an input image and produce style tags, with a live example producing 1980s hardcore techno output complete with female vocalist tags.
The tutorial also briefly covers the Gradio web interface as an alternative to ComfyUI. Viewers interested in open-source audio generation, local AI workflows, or replacing cloud music tools will find concrete, reproducible steps alongside live audio demonstrations of the model’s output quality.
📺 Source: Nerdy Rodent · Published February 04, 2026
🏷️ Format: Tutorial Demo
