What the AI Music Extender Actually Does
Every run starts at a second you choose. The AI music extender reads the audio ahead of that point — key, tempo, chord movement, which instruments carry the part — and writes new bars forward from there. Nothing is copied back from earlier in the song, so a chorus does not simply return. The rest you steer in words: a style line holds the direction or asks for a change, an exclude list keeps unwanted genres out, and three sliders decide how literally the model takes you. Vocals carry on with lyrics you write, or drop out entirely when the section has to sit under a voice-over. Source audio comes from your library or from a file you upload as MP3, WAV, M4A or OGG, under eight minutes and 50MB. A run lands in about one to two minutes with the credit cost on screen first, so a second attempt stays a small decision.