ElevenLabs unveils enhanced music AI model with global access

ElevenLabs, an AI voice generation company, launched its latest generative music model, ElevenMusic v2.5, on September 3, 2026, making it available globally through apps and APIs. The update improves audio richness and naturalness, with most listeners preferring v2.5 in blind tests involving 47,885 sample pairs, especially in R&B, soul, hip-hop, rock, and orchestral genres. The platform offers free and professional tiers for v2.5, with free users getting 5 lossless downloads daily and professionals receiving 400 monthly downloads. Users retain ownership of generated tracks but are restricted from imitating specific artists or using copyrighted material. ElevenLabs emphasized that v2.5 was trained on a legally authorized dataset, distinguishing it from competitors like Suno, which faced recent copyright lawsuits. The company clarified that its previous agreement with Universal Music Group (UMG) does not apply to v2.5, highlighting the growing importance of copyright compliance in the generative audio industry.