Suno has rolled out version 5.5 of its AI music model, marking one of its most significant updates to date. While previous releases focused on improving sound quality and generating more natural vocals, this update shifts the focus toward user control and personalization.
The latest version introduces three key features: Voices, My Taste, and Custom Models, all designed to give creators more influence over how music is generated and performed.
One of the headline additions is Voices, a feature that allows users to train the AI using their own voice. Users can upload acapella recordings, full tracks, or record directly through a microphone. According to Suno, higher-quality recordings reduce the amount of training data needed. To address potential misuse, the platform requires users to record a verification phrase during setup, although concerns remain about whether advanced voice cloning tools could bypass such safeguards.
Once trained, the AI can replicate the user’s voice across new compositions or apply it to existing music, effectively allowing users to “sing” on tracks without recording new vocals.
Custom Models takes personalization further by allowing users to train the system on their own music catalog. Users must upload at least six songs to create a model that reflects their style. The trained model can then guide how Suno interprets prompts, helping generate music that aligns more closely with a user’s unique sound.
The third feature, My Taste, focuses on passive personalization. It tracks user behavior over time, including preferred genres, moods, and styles, and uses that data to influence future outputs. When generating music using Suno’s automated tools, the system applies these learned preferences to produce more relevant results.
Suno says My Taste will be available to all users, while Voices and Custom Models will be limited to Pro and Premier subscription tiers.
With v5.5, Suno is moving beyond basic AI music generation toward a more customized creative experience. By combining voice replication, style training, and behavioral learning, the platform is positioning itself as a more powerful tool for both casual creators and professional musicians exploring AI-assisted workflows.
