Voice AI company Celebrities has announced a new version of its music generation model called Music v2, which can switch genres mid-track. The company says the model is designed to handle both vocal and compositional complexities. This new release comes nearly 10 months after the startup unveiled the first version of its music generation model.
Eleven Labs noted that this model could transition from opera to heavy metal and vice versa, provide fast raps without losing consistency, and add non-musical sound effects to tracks. The new model allows artists to select a part of a song and use prompts to recreate it without touching other parts of the track.
Additionally, instead of generating short clips, artists can build songs into sections, such as intros, verses, and choruses, and stitch them together. ElevenLab added that this model delivers more reliable performance across language, lyrics, vocals, and arrangements.
For the past few months, AI Labs has been racing to release models that can generate professional music. Google, Stability AI, and Suno have also released new music generation models with the ability to generate longer and more complex tracks. At the Google I/O developer conference, Google added the ability to easily create covers, edit songs in sections, and generate music videos using the Flow Music tool.
Eleven Labs emphasized that the new model is built on licensed data and allowed for commercial use, so users are free to use the truck as they see fit. The impressive deals with labels are key, considering other AI music startups like Suno and Udio are facing lawsuits over copyright issues.
This new model is available in Eleven Lab’s Eleven Creative tools for marketing and branding teams, the newly launched Eleven Music platform for creating AI-generated songs, and will soon be available in the Eleven API.
If you buy through links in our articles, we may earn a small commission. This does not affect editorial independence.
