Since launching in 2023, AI music platform Suno has received a steady torrent of criticism from musicians and music fans sceptical of its legally dubious training practices and wary of the growing threat that generative AI poses to the creative industries.
Despite this – and a mounting industry backlash that’s seen multiple lawsuits filed against the company – Suno has grown to serve more than 2 million paid subscribers and raised hundreds of millions in capital investment.
However, it seems that with the release of its latest V6 model, some of Suno’s most enthusiastic supporters are abandoning ship. The response from the platform’s user community has been overwhelmingly negative, with many claiming to have cancelled their subscriptions as a result of the changeover.
Latest Videos FromMusicRadar
Suno unveiled V6 earlier this month, a new suite of music generation models developed in partnership with major record labels following licensing deals inked with Warner Music Group, BMG and Believe. Suno claims that the V6 models are its best yet, describing them as “faster, more expressive and higher quality” with a more nuanced comprehension of musical language.
Much of the criticism levelled at V6 has centred on its audio quality. One user on Suno’s Reddit page described the results as “muffled, dull and strangely lifeless”, adding that “the vocals are often buried, the high end feels almost completely gone, and the whole mix has that ‘blanket over the speakers’ feeling.”
“I’m giving them a month to fix this problem and if not, I am cancelling my subscription,” another responded in the comments.
Others have found fault with the “creativity” of V6’s output, observing that the music it produces has become more generic and bland, with one user describing the results as “flatter, soulless and more similar-sounding”. Another described “overproduced, heavily compressed and limited tracks that lack even a semblance of melody, grit or soul”.
“I’ve published two albums made with V5.5 – people tell me the songs, vocals and instrumentation are virtually indistinguishable from ‘real’ music,” one particularly disgruntled user wrote. “I spent the two days of free usage trying all the same styles and prompts in V6 and found it produced muffled, generic, soulless garbage by comparison.”
“I haven’t been able to generate a single track I like,” another complained. “It completely lacks soul and basic understanding of how to compose a track, everything feels so random.”
Many of those that are unhappy with V6 have been calling for Suno to bring back older models that have now been permanently retired since the V6 launch, likely due to the increasing pressure from major record labels and industry players. Despite having settled with WMG, Suno remains embroiled in several long-running legal disputes, including with Universal Music Group and Sony Music Entertainment.
Some users are speculating that the drop in quality and creativity they can hear from V6 is due to a drastically restricted training dataset. Where earlier models were trained on tens of millions of recordings – in Suno’s words, “all music files of reasonable quality that are accessible on the open internet” – V6 has been limited to licensed material from the catalogues of its industry partners and user data.
While the exact scope of V6’s training data hasn’t been publicly disclosed, it’s undoubtedly smaller and less diverse than the dataset used to train previous models.
Suno’s move towards licensed training data represents a welcome development for musicians and music fans that were uncomfortable with the company’s training practices, but if the negative response to V6 persists, the company may find that its decision to legitimise those practices ends up driving away the very users who helped fuel its rise. For those of us who remain sceptical about what AI music means for musicians and the wider industry, that may not be such a bad thing.
