AI music generator Suno has launched its v6 family of models, marking the first time it has trained new systems using data licensed from record labels, including Warner Music and BMG. The launch is significant because it formalises Suno's commercial relationship with major labels following a 2024 copyright infringement lawsuit, with Warner dropping its case last year in exchange for a licensing deal giving artists control over how their voices and music are used. Suno says the partnership will generate revenue for participating labels and artists from the day of launch.

The v6 series comprises three models: the flagship v6, described as "reliable, precise" and consistent across genres; v6-wild, a more experimental model aimed at helping musicians brainstorm; and v6-mini, the only one available to free users, with the other two requiring a paid subscription. New features include natural-language editing of song sections, mashups combining elements from multiple tracks, more precise sampling, and the ability to generate music from audio, images or video, as well as replace individual lyrics without regenerating a whole song. Suno's chief product officer, Jack Brody, said the models were trained on a new dataset including licensed and user data, though questions remain over the company's past practices after a 2026 leak revealed it had scraped music and podcasts from platforms including YouTube and Deezer to train earlier models, which are now being retired.