Suno launches v6 music-generation models, retiring all previous versions

Suno launched v6 on September 9, 2026, a new generation of its AI music models built with input from major record labels, and retired every earlier Suno model in the process.

What's new

Suno describes the new models as "faster, more expressive and higher quality, with powerful new features that give creators greater control" over the music they generate. The release comes in three variants aimed at different use cases. The flagship v6 model targets "reliability and precision across all genres" and is available to Pro and Premier subscribers. v6-wild is built for exploration, producing "unexpected, textured and ambitious results," also gated to Pro/Premier. v6-mini is the free-tier option, a faster and lighter model Suno says delivers "better results than existing free platforms" and is available on every plan.

On the input side, v6 can take a written idea, a voice memo, an image, or a video and turn it into a finished song, and it can read the mood of a reference track and use that feeling as a starting point for something original rather than a direct copy. It also adds targeted editing: users can change a single lyric line without regenerating the rest of the song.

This is Suno's first model family developed directly with record labels, with Warner Music Group, BMG, and Believe named as industry partners on the release.

Context

Suno has faced ongoing legal and industry pressure over how its earlier models were trained, including lawsuits from major labels alleging unauthorized use of copyrighted recordings. The v6 launch — built with label partnerships rather than against label objections — marks a shift toward licensed, industry-cooperative model development, following on Universal Music Group's separate licensing deal with ElevenLabs earlier in September. Retiring all prior Suno models simultaneously with v6's release is also a bigger step than a typical version bump; it forces the platform's entire user base onto the new, label-partnered generation at once rather than letting older models run in parallel.

Why it matters

AI music generation has been one of the most legally contested corners of generative AI, with labels treating unlicensed training data as an existential threat to catalog value. A flagship AI music platform shipping its next generation of models with named label partnerships, rather than continuing to operate adversarially, is a meaningful signal that the standoff between AI music tools and the music industry is starting to resolve into licensing deals rather than pure litigation — a path other AI audio and video generators working with copyrighted training data may increasingly be pushed toward.