Suno, a leading generative artificial intelligence music creation program, officially rolled out its highly anticipated v5 model in beta on September 23, 2025, making it available to Pro and Premier subscribers. The new iteration promises significant advancements in AI-generated music, though one user, identified as Andy, noted it is "still several iterations from musical AGI," despite being a "very impressive model." Andy, who tests models with challenging "orchestral djent" prompts, highlighted the model's progress.
Suno's official announcement claims that v5 delivers "more immersive audio, authentic vocals, and unparalleled creative control," positioning it as the "world’s best music model." According to Suno CTO Georg Kucsko, v5 represents a "big step forward in core technology" that will also power upcoming platform features. An ELO benchmark further supports these claims, with v5 scoring 1,293, surpassing v4.5+ (1,208) and v4 (992).
Early feedback from users, including Vsévolod, lead singer of Singular Decomposition, points to a "colossal gap" in vocal quality compared to previous versions, with enhanced expression and correct stress placement. However, Vsévolod observed that instrumental quality, particularly electric guitars, still sounds "dirty and blend together," indicating room for improvement in complex genre blending. The model excels with similar genres but struggles with combining disparate styles like "Italian opera and Californian trap."
v5 model is set to be the foundation for future tools, including the upcoming Suno Studio, an audio workstation designed to offer granular control over song structure. Suno Studio will allow users to "restructure songs by removing and adding individual components like drums, synthesizers, and vocals." While currently exclusive to paid subscribers, Suno has indicated that a significant upgrade for free users is also in the pipeline.
The launch comes as Suno, alongside rival Udio, faces ongoing copyright infringement lawsuits from the Recording Industry Association of America (RIAA), which alleges the models were trained on copyrighted music. Suno maintains that its use of copyrighted material is transformative. The company continues to push the boundaries of AI music generation, aiming for broader creative utility despite existing challenges.