Artificial intelligence voice generation startup ElevenLabs officially launched its new ElevenMusic Music v2.5 version on the 3rd, and simultaneously made it available worldwide to developers and users through mobile applications and APIs. As the latest iteration in the field of generative music, Music v2.5 further enhances the richness and naturalness of the audio. In a blind test comparison based on 47,885 pairs of samples, most listeners showed a clear preference for the v2.5 version, particularly excelling in mainstream and complex genres such as R&B, soul, hip-hop, rock, and orchestral music.

In terms of commercialization implementation, the product continues to retain the previous v2 version, while offering free and professional (Pro) tiers for Music v2.5. Free users can enjoy 5 lossless downloads per day, but commercial use requires proper attribution. The professional tier offers 400 downloads per month. The platform grants users full ownership of generated tracks, but strictly limits the imitation of specific existing artists' voices in usage guidelines and prohibits downloading tracks derived from others' copyrighted works.

Facing frequent copyright disputes caused by generative AI, ElevenLabs emphasized that Music v2.5 was trained on a legally authorized music dataset, thereby distinguishing itself from competing products like Suno, which have recently been involved in copyright lawsuits.

Although the company had previously reached a strategic licensing agreement with Universal Music Group (UMG), the official confirmed that this agreement serves future independent products and was not used for the current Music v2.5. With competitors like Suno updating their models, the release of Music v2.5 not only marks a key step forward in AI music generation across multiple genres, but also highlights that copyright compliance is gradually becoming a core commercial barrier in the generative audio industry.