AI music startup Suno has released version 4 of its music generation model, bringing improvements to audio quality, lyric accuracy, and song structure.
The v4 update introduces "Remaster," a feature that lets users upgrade tracks made with older versions of Suno to v4 quality. The company has also added "ReMi," an AI-powered lyrics assistant designed to help users with songwriting. v4 is currently available in beta for Pro and Premier subscribers, with no timeline announced for free users.
Sample song generated with Suno v4 | Video: Suno
In addition, Suno has overhauled its cover art creation system, saying it now creates more distinctive designs that better match specific music styles. The company has also updated existing features such as "Covers," which allows users to reinterpret uploaded audio files, and "Personas," which saves style settings for future use.
Early user feedback varies
Early user feedback on Suno's Discord channel shows mixed results. While many users report clearer, more accurate lyrics, others point to technical issues. Some users have noticed problems with cymbal artifacts, reverb effects, and insufficient volume on metal and rock tracks.
One user, Mook, noted that the new remastering feature is a "big hit or miss." Other users say that while the audio quality improves with remastering, songs sometimes lose their original character.
The updated Personas feature is also showing some early issues, with users reporting that it tends to repeat lyrics from original songs instead of properly implementing new ones.
The v4 release comes as both Suno and competitor Udio face legal challenges from music companies. The lawsuits allege that these companies trained their AI models on copyrighted music without proper permission.
