ElevenLabs has launched Music v2.0, a cutting-edge music generation model featuring significant upgrades in vocal synthesis and instrumental accompaniment. The new model introduces localized inpainting and segmented structured composition, allowing creators to regenerate specific audio segments without altering the entire track. This innovation supports explicit definition of song sections, enhancing the structural integrity of long-form audio. Music v2.0 also excels in acoustic performance, enabling high-density lyrics and rapid rap delivery, along with seamless genre transitions within a single track. The model is trained on legally licensed data, offering full commercial usage rights and eliminating licensing fees for content creators. Additionally, ElevenLabs has reduced pricing for its services, cutting API rates by 50% and Creative subscription fees by 40%.