ElevenLabs Releases Music v2.5 with App and API Access

ElevenLabs has officially released Music v2.5, the latest iteration of its AI music generation model, making it accessible through both its web application and developer API. The company designed this update to deliver fuller, more natural-sounding audio. In a blind test involving 47,885 comparison pairs, listeners consistently preferred Music v2.5 over previous versions, with the most notable improvements occurring in genres such as R&B, Soul, Hip-Hop, Rock, and orchestral music. The older v2 model remains accessible for users who still require it.

The model is available under two main tiers. The free tier allows users to make five lossless downloads per day, though commercial use under this option requires attribution. Meanwhile, the Pro tier offers 400 lossless downloads per month. Regardless of the tier, users retain the rights to their generated tracks. However, ElevenLabs has implemented strict guardrails to prevent copyright issues: the system blocks downloads of tracks based on other artists' songs and explicitly prohibits the imitation of existing musicians.

For developers and creators, the release provides a legally safer alternative to other AI music tools. ElevenLabs trained its existing music models on licensed stems and music, contrasting with competitors like Suno, which currently faces lawsuits over unauthorized training data. Although ElevenLabs recently signed a licensing agreement with Universal Music Group, the company clarified that this deal is reserved for future, separate products and does not apply to Music v2.5.

For audio engineers, app developers, and content creators, the integration of Music v2.5 into the ElevenLabs API opens up new avenues for automated, high-quality soundtrack generation. By combining improved acoustic fidelity with a clear licensing framework, the update allows practitioners to confidently integrate AI-generated music into commercial applications without the looming threat of copyright infringement.