AI music generator Suno has released its v6 model family, marking its first release developed with support and licensed content from record industry partners including Warner Music Group, BMG, and Believe. The release introduces three distinct variants—v6, v6-wild, and v6-mini—trained on a new dataset separate from previous generations, alongside user-provided data.
The update adds significant functional upgrades, including natural language chat editing for specific track components like lyrics or guitar lines without full regeneration, multi-element track mashups, and multimodal input support for images, video, and audio. While the model shows improved understanding of complex music genres, tests indicate it remains incapable of producing intentional dissonance or non-harmonic imperfections, continuing to generate rhythmically and harmonically polished outputs.
Suno has begun rolling out v6 today and announced plans to eventually phase out its older model iterations.
Why it matters
Licensing deals with major labels mitigate copyright risks, offering a blueprint for compliant generative media platforms.
In-context natural language editing and multimodal prompting significantly lower friction for audio and music workflows.
Source: theverge.com
