Amuse AI setup: local music and images on Windows

This Italian-language tutorial covers the standalone Windows setup of Amuse AI 3.6.1 and demonstrates music and image generation. The presenter extracts the release ZIP into a folder on drive D to save space on the system drive, then opens the application there. Drive C is also an option. He describes generation as offline after setup, but the walkthrough requires model downloads and a Hugging Face account for access tokens used by some models.

For music, he downloads ACE Step 1.5 XL and lets the application create a Python virtual environment. He uses ChatGPT to prepare separate style and lyrics prompts for English and Italian songs, then saves the audio. The demonstrated controls include BPM, duration and reference audio. His two generations take about a minute each; these are results from his computer, not general performance guarantees.

The image section uses Z-Image and FLUX.1 Schnell to run models locally. It covers loading downloaded models, choosing image dimensions, entering negative prompts and saving PNG files. A read token goes into Amuse AI's general settings. The presenter notes that a landscape result still resembles an illustration despite a negative prompt asking it to avoid cartoons.

Video generation is deferred to another tutorial. He reports an earlier LTX 2.3 Distilled test on a 12 GB Nvidia graphics card that took roughly half an hour, so this walkthrough does not demonstrate the video workflow.