ElevenLabs Launches AI Music Model, Expands Beyond Voice
ElevenLabs, the artificial intelligence audio-generation unicorn renowned for its pioneering text-to-speech technology, has officially ventured into the burgeoning realm of AI music creation with the launch of its new model, Eleven Music. This strategic expansion marks the company’s first significant move beyond its established voice synthesis, conversational bots, and speech translation tools since its founding three years ago. The new offering distinguishes itself by enabling users to generate music that is explicitly cleared for commercial use, addressing a critical concern plaguing the nascent AI music industry.
Central to Eleven Music’s debut are groundbreaking licensing agreements with key players in the music industry. ElevenLabs has forged partnerships with Merlin, a global digital rights agency for the independent music sector, and Kobalt Music Group, one of the world’s largest music publishers. These collaborations are designed to establish a transparent and rights-protective framework for AI-generated music, a stark contrast to the legal challenges faced by other AI music startups like Suno and Udio, which have been subject to lawsuits from the Recording Industry Association of America (RIAA) over alleged copyright infringement from unlicensed training data.
Under these agreements, artists and songwriters represented by Merlin and Kobalt can voluntarily opt-in to have their works used for training ElevenLabs’ AI models. This opt-in mechanism is coupled with a revenue-sharing model, ensuring that participating artists directly benefit from the use of their music. Notably, Kobalt’s agreement includes a “Most Favored Nation” clause and aims for “parity” with recorded music revenues, proposing an approximate 50/50 split of royalties generated from the AI platform. A spokesperson for ElevenLabs affirmed that “Eleven Music was built in partnership with artists, labels, and publishers, and includes guardrails to protect rightsholders,” with the model strictly trained on licensed data.
Dubbed Eleven Music v1.1, the new model offers robust capabilities for creators. Users can generate studio-grade music from simple natural language prompts, with comprehensive control over genre, style, and structure. The platform supports both instrumental and vocal formats, offering multilingual vocals in English, German, Spanish, and Japanese. It introduces two modes of operation: a Simple Generation mode for rapid song creation and an Advanced Section-Based Generation mode, allowing users to define individual song sections—such as intro, verse, chorus, and bridge—with specific attributes like tempo, tone, and lyrical structure. Additionally, the v1.1 update features real-time streaming during generation, enabling users to preview finished portions while subsequent sections are still rendering, significantly accelerating the iteration process. This makes it a powerful tool for diverse commercial applications, including film, television, games, podcasts, advertisements, and social media content.
ElevenLabs, valued at $3.3 billion following a $180 million Series C funding round in January, has been testing the platform with approximately 20 clients who have already utilized the model to create content for various media. The company’s proactive approach to licensing positions it as a leader in navigating the complex legal and ethical landscape of AI music, which has seen rapid growth and is projected to reach a market size of around $500 million in 2025. While concerns about copyright and the ethical implications of AI-generated music persist, ElevenLabs’ strategy aims to foster collaboration rather than conflict, charting a course for responsible innovation in the evolving music industry.
[ElevenLabs just dropped AI music that’s actually legal, setting a new harmony for creators and the industry.]