Suno has launched a new generation of music models built with Warner Music Group, BMG and Believe, and is switching off the models trained on material the labels said was used without permission.
The v6 lineup includes three models: v6 for more polished results and v6-wild for less predictable output, both for Pro and Premier subscribers, and v6-mini for everyone. They support partial song editing, mashups from multiple sources, audio isolation and sampling, and users can prompt them with text, audio, images or video.
The release follows through on a commitment Warner extracted when it settled its copyright lawsuit against Suno in November 2025: the agreement required Suno to launch more advanced licensed models and retire its existing ones. Suno still faces suits from labels like Sony and Universal Music and from artists including Jason Isbell, and the company recently admitted to training earlier models on YouTube videos. It has also said it plans to watermark generated songs.
Artists working with Believe can opt into the new Suno experiences, and songs created on Suno can become eligible for distribution through Believe and TuneCore, giving AI-generated music a route into mainstream stores. Suno says little about what the new models were trained on, naming its partners as co-developers without detailing catalogs or volume. For the AI music sector, this is the first big test of whether licensed training data can quiet the copyright wars.
