ElevenLabs has unveiled a generative AI-based music production model. The model features improved genre switching and vocal expression, characterized by enhanced editing and control capabilities during the music generation process.

Eleven Labs announced on the 22nd that it has released its AI music generation model, 'Music v2'. The key feature of the model is that it maintains the capabilities of the previous version while improving vocal expression and arrangement quality.

Music v2 is designed to allow for a seamless transition between different genres within a single song. For example, different genres such as opera and heavy metal can be arranged consecutively on a single track.

It also includes the ability to reliably generate fast rap vocals and complex rhythm patterns. Another feature is the ability to combine non-musical sound effects (SFX) with background music.

AI music generation models are technologies that automatically generate melodies and vocals based on input text or sound sources.

Music v2 provides an editing function that allows you to modify only specific sections during the music creation process. You can reconstruct only parts of the song without regenerating the entire song.

Key features include: △style analysis generation using reference audio sources △specific section regeneration function △support for multilingual vocal generation △genre switching and complex vocal expression △step-by-step generation based on song structure.

This model is provided through three platforms: Eleven Music, Eleven API, and Eleven Creative.

The company stated that it built the model based exclusively on licensed data. It also explained that it improved accessibility by lowering usage fees for APIs and content creation platforms.

Eleven Labs announced that it plans to expand support for customized models and music sharing features through future feature expansions.