ElevenLabs releases Music v2 with mid-track genre switching and commercial licensing
The latest iteration of ElevenLabs’ music generation tool allows for section-specific regeneration and complex vocal handling, positioning the startup against competitors in a market increasingly focused on copyright compliance.
Voice artificial intelligence company ElevenLabs has launched Music v2, its latest music generation model, which introduces the capability to switch genres mid-track. The release marks a significant update to the firm’s audio technology, arriving nearly 10 months after the startup initially introduced its first music generation model. The new system is designed to handle complex compositions, including the ability to transition from opera to heavy metal and back, as well as delivering coherent fast rap and integrating non-musical sound effects.
A key technical advancement in Music v2 is its capacity for section-specific regeneration. Users can now select specific parts of a song, such as an intro, verse, or chorus, and recreate them using text prompts without altering the remainder of the track. This functionality allows artists to construct songs by building and stitching together distinct sections rather than generating short, disconnected clips. ElevenLabs stated that the model also performs more reliably across different languages, lyrics, vocals, and musical arrangements compared to previous versions.
The company emphasised that Music v2 is trained on licensed data and is cleared for commercial use. This distinction is a strategic differentiator in an AI music sector where copyright concerns have become a central issue. Competitors such as Suno and Udio have recently faced court cases regarding copyright infringement, making licensed training data a critical factor for enterprise adoption. ElevenLabs noted that securing deals with labels is essential for long-term viability in this space.
The launch occurs amid intensifying competition from major technology firms and other startups. Google recently announced Flow Music at its I/O developer conference, offering tools to create covers, edit songs by sections, and generate music videos. Similarly, Stability AI and Suno have released new models capable of producing longer and more complex tracks. The market is currently defined by a race to deliver professional-grade music generation capabilities while navigating the legal complexities of intellectual property.
Music v2 is currently available through ElevenLabs’ ElevenCreative tool, which targets marketing and branding teams, as well as via the newly launched ElevenMusic platform designed for creating AI-generated songs. The company indicated that access to the model via the ElevenAPI is coming soon, though no specific timeline was provided. This rollout positions ElevenLabs to capture demand from both creative professionals and commercial entities seeking compliant AI audio solutions.
