Suno is betting big on legitimacy. On Wednesday, the AI music startup unveiled Suno v6, a new model family built entirely on licensed data from major industry players like Warner Music Group, BMG, and Believe. This marks a sharp pivot from its earlier days, where the company faced multiple lawsuits for allegedly scraping copyrighted material without permission. Suno settled with Warner last year and closed a deal with BMG just last month, signaling that the path forward involves paying for training data rather than fighting over it in court.
The new lineup offers three distinct tiers. Paying users get access to the base Suno v6, designed for reliable and steerable outputs, plus Suno v6-wild, an experimental variant meant for unpredictable ideation. Everyone else can use Suno v6-mini, a faster, lightweight version. The company is retiring its older models, pushing users toward this new infrastructure. Functionally, v6 allows for granular control: you can edit specific song parts via text prompts, use images or video as reference points, and even isolate instruments from samples to build new beats. It is a significant step up in precision compared to previous generations.
Jack Brody, Suno’s chief product officer, frames these updates as revenue engines. He told TechCrunch that derivative works create new income streams for rights holders and artists, provided they opt into Suno’s upcoming remix program. The company also plans to introduce watermarks and has already tightened download limits to curb low-intent mass exporting. Brody argues that while Suno can help manage streaming fraud, the ultimate governance of content on distribution platforms remains with those distributors.
Yet, the legal cloud hasn’t fully cleared. Suno still faces active lawsuits from Sony, Universal Music Group, and artist Jason Isbell. Complicating matters, the company admitted just a day before this launch that it trained earlier models using YouTube videos. With over $819 million in funding according to PitchBook, Suno has the capital to keep building, but whether licensed data alone satisfies the broader creative community remains an open question.
