Back to News
Tools

Suno ships v6 music models with Warner, BMG, Believe deals as labels' lawsuits grind on

Suno ships v6 music models with Warner, BMG, Believe deals as labels' lawsuits grind on
Image: The Decoder

Suno's new v6 AI music generator, built alongside Warner, BMG, and Believe, replaces all older models — but the company still won't disclose what went into training.

September 9, 2026 · 3 min read · VintuxAI Newsroom

Suno released its sixth-generation music models on Wednesday, a three-version lineup developed in partnership with Warner Music Group, BMG, and Believe. The company simultaneously shut down all previous models, meaning users now work exclusively with v6. It is the clearest signal yet that Suno is betting its future on licensed collaborations rather than the freewheeling training data approach that got it sued by the industry's biggest labels.

The new models come in three tiers: a standard version, one optimized for vocal clarity, and another aimed at instrumental composition. Suno says users can now edit parts of a finished song with text commands — tweaking a chorus or swapping an arrangement without regenerating the whole track — and can generate music multimodally by feeding in text, audio, or even images as starting points. Those features push Suno closer to a full production tool rather than a novelty generator, though the company declined to specify which catalogs or artists were used to train v6.

That silence matters because Universal Music Group and Sony Music have active copyright infringement lawsuits against Suno, claiming its earlier models were trained on their recordings without permission. Warner, BMG, and Believe have chosen the opposite path, striking licensing deals that give Suno access to their repertoires and revenue-sharing arrangements in exchange for legitimacy. The contrast has split the industry: some labels see Suno as a threat to be crushed in court, others as a partner to monetize before the technology outruns them.

Suno's partnership with Believe is particularly notable. Believe is a major distributor for independent artists and labels, and its involvement suggests Suno is courting the mid-tier and indie sector that often gets squeezed when big labels set terms. Warner's deal, meanwhile, was already announced in May, but the v6 launch marks the first time Suno has shipped models explicitly built under those agreements. The company says older models were retired to ensure all output now runs through the licensed pipeline, a technical choice with legal implications.

For users, the practical shift is immediate: any saved projects or shared tracks created on older Suno models will no longer regenerate or edit, though existing audio files remain playable. The company's move to consolidate on v6 also simplifies its product surface — no more choosing between model versions — but it raises the stakes on whether the new models match the creative range of the ones they replaced. Early demos emphasize cleaner vocals and more precise instrumental separation, though Suno has not published independent benchmarks.

The open question is whether Universal and Sony will see v6 as a provocation or a negotiating lever. Suno's licensing deals give it cover for the new models, but the lawsuits target the old training data, and the company has not conceded wrongdoing in either case. As the music industry watches v6's reception, the real test may be whether Suno can prove that licensed AI music is commercially viable — or whether the labels suing it are simply waiting for the startup to run out of runway.

What to watch next: whether v6's multimodal generation and partial editing features draw enough paying subscribers to justify the licensing costs, and whether Universal or Sony shift their stance if Suno's partnership model starts showing real revenue. A court ruling on the older training data could still upend everything, but v6 gives Suno its best argument yet that the future of AI music is being built with the industry, not against it.

Written in-house by the VintuxAI newsroom, in our own words. The story was first reported by The Decoder.