AI voice and music specialist ElevenLabs has released Music v2, an upgraded music generation model that lets creators build songs section by section and seamlessly switch genres within a single track. Launched roughly 10 months after the company’s original music model, Music v2 expands on AI audio capabilities and gives artists and creators finer control over structure, vocals, and stylistic transitions without losing musical coherence.
The most striking ability of the new model is its capacity to move fluidly between genres, for example, from opera to heavy metal to fast rap, within one composition while keeping vocal quality, rhythm, and instrumentation aligned. Creators can instruct the system to generate or regenerate specific parts of a song, such as intros, verses, or choruses, and stitch sections together into a full piece just as they would in traditional music production.
Music v2 isn’t just about longer tracks; it represents a shift toward section‑based song construction. Artists can define a song’s structure using natural language prompts, then regenerate only chosen sections without affecting the rest of the track. The model also supports complex vocal arrangements, multilingual lyrics, and compositional nuances that earlier AI systems often struggled to maintain.
This marks a step beyond short, loopable clips toward more complete, studio‑style output. By enabling genre transitions mid‑track and contextual understanding of arrangements, the model brings AI music generation closer to professional production workflows.
Music v2 powers multiple ElevenLabs products:
All tracks generated with Music v2 are built on licensed data and cleared for commercial use, a notable distinction from some earlier AI music tools that faced legal challenges around training data and copyright.
The release of Music v2 comes amid intense competition in AI‑generated audio. Other labs, such as Stability AI, have recently launched models that produce longer and more complex compositions, and major tech companies like Google have added generative music tools to their creative toolkits.
ElevenLabs’ genre‑hopping capability and section‑by‑section control address two persistent limitations of earlier models: a lack of coherent long‑form structure and rigid stylistic boundaries. This opens doors to applications ranging from soundtrack creation and advertising jingles to experimental compositions and rapid prototyping for producers.
The combination of genre versatility, sectional editing, and commercial readiness positions Music v2 as a powerful tool for both creators and businesses. Its availability across ElevenLabs’ suite from creative platforms to developer APIs, suggests growing confidence in AI music as a practical, extensible technology rather than a novelty.
As the AI music space evolves, models like Music v2 signal that the next phase will involve not just automatic generation, but fine‑grained control and integration into existing creative workflows turning speculative demos into tools that professionals are more likely to use in real projects.
When Unlucid AI Is Useful, and When It May Not Be EnoughRoughly four o...
My younger brother was doing exactly what most students do two weeks b...
OpenAI is still actively developing its long‑mooted “super app”, a sin...
I started looking for Wsup AI alternatives after a simple realization:...
Last week, I had a small pile of visual jobs sitting in a folder, none...
Marvell Technology is set to join the prestigious S&P 500 index la...
Discussion