Photo credit
Generative AI music company Suno has raised $125 million in its latest funding round, according to a company blog post. One of the few startups that can simultaneously generate voice, lyrics, and instruments, the AI music company wants to usher in “a future where anyone can make music.”
Suno allows users to create complete songs from simple text prompts. Most of the company's technology is proprietary, but it does use OpenAI's ChatGPT for lyric and title generation. Free users can create up to 10 songs per month, while the Pro ($8/month) and Premier ($24/month) plans allow users to create up to 500 or 2,000 songs per month, respectively, and are subject to “general commercial terms.”
In its announcement, the company cited investors including Lightspeed Venture Partners, Nat Friedman and Daniel Gross, Matrix and Founder Collective. Suno also said it has worked closely with a team of advisors including 3LAU, Aaron Levie, Alexandr Wang, Amjad Masad, Andrej Karpathy, Aravind Srinivas, Brendan Iribe, Flosstradamus, Fred Ehrsam, Guillermo Rauch and Shane Mac.
Suno is widely considered to be one of the most advanced AI music models on the market today, but in past interviews the company has not disclosed what material its training data includes. Expert Ed Newton-Rex, founder of Fairly Trained and former VP of Audio at Stability AI, warned in a recent article in Music Business Worldwide that given how Suno was able to generate music using a model that closely resembles copyrighted material, it was likely trained on copyrighted material without consent.
In a recent Rolling Stone article about the company, investor Antonio Rodriguez said that Suno's lack of licensing deals with music companies isn't a concern, saying the lack of licensing deals “is a risk we had to take on when we invested in the company, because we're the next big thing to get sued by these companies. … To be honest, if this company had signed with a label when it started, I probably wouldn't have invested in it. I think they had to make this product without constraints.”
However, Suno representatives have previously said that their model doesn't allow users to create music using prompts like “Radiohead-style ballad” or to adopt the voice of a specific artist.
Many AI companies, including OpenAI, argue that training on unlicensed copyrights is “fair use,” but the legality of the practice has yet to be determined in the U.S. The New York Times has sued OpenAI for training on copyrighted archives without consent, credit, or compensation, and music publishers including Universal Music Group, Concord, and ABKCO have sued Anthropic for using song lyrics to train its large-scale language models.
In a Suno blog post, CEO Mikey Shulman wrote: “Today, we're excited to announce that we've raised $125 million to build a future for music where technology amplifies rather than replaces our most precious resource: human creativity.”
“We released our first product eight months ago, empowering anyone with just a simple idea to make a song,” he continued. “We're still early days, but we're already seeing 10 million people making music with Suno. We have Grammy-winning artists using Suno, but our core user base is made up of everyday people who are often making music for the first time.”
“We've watched producers dig records, friends swap memes, and streamers co-create songs with stadium-sized audiences. We've helped artists who lost their voices reinvent lyrics after decades on the sidelines. We've seen teachers spark their students' imaginations by transforming lessons into lyrics and stories into songs. And just last weekend, we heard heartwarming stories of mothers moved to tears by songs their loved ones created with Suno's help.”