MUSI-CO

MUSI-CO is a Florence-based music AI company whose engine converts video, images, text or emotional cues into structured MIDI compositions. It is built for composers, game and media studios who want editable notes rather than finished audio files.

The flagship ANY2MUSIC engine reads the structure and mood of a source medium and maps it onto musical parameters such as harmonic tension, density and rhythm. Output is MIDI, so you can change instruments, edit notes and velocities, and continue the work in a DAW. Its models are trained on datasets written by hired composers rather than scraped music.

Other tools include Syntagma for evolving generative loops in scoring, games and installations, LiveAI for low-latency accompaniment on stage, and Listen for analyzing compositions. Processing can run fully offline on your own hardware, and consumer iOS apps Mindkestra and MusicFit are built on the same engine.