Suno v5.5

Suno v5.5, announced by co-founder and chief executive Mikey Shulman on 26 March 2026, is the company's current text-to-music model. Like every version since V3 it turns a written description into a finished song — vocals, lyrics, arrangement and mix — but the release was not sold on audio length or fidelity. It was sold on personalisation, and it is the point where the product stopped being a generic music machine. Three features define it. Voices lets a Pro or Premier subscriber register their own singing voice: the user records or uploads audio, and a verification step matches that singing voice against a randomly chosen phrase, so the voice being registered has to be present at the microphone. Registered voices are private — Suno states that only their owner can use them to generate songs, with sharing promised as a later, user-controlled option. Custom Models let the same subscribers train up to three personal models on tracks from their own catalogue, so the output follows their style rather than a genre average. My Taste, available on every tier including free accounts, learns which genres and moods a listener keeps returning to and biases generations towards them. Suno calls it "our best and most expressive model yet". The company has not published length limits, benchmark scores or a technical description of the architecture, and wujec.ai does not invent them: what is documented is the feature set and the tier split. The unresolved training-data litigation between Suno and the recorded-music industry applies to this version exactly as it did to v5 — and Voices adds a second, separate question, because a verified singing voice is biometric-grade data held by a company whose data practices are already in court. Suno v5 (September 2025) keeps its own profile in this catalogue; new versions never replace old entries here.

▸Videos

No videos yet.

▸Reviews