Suno released v5.5 on March 26, bringing voice cloning, custom models, and personalized taste profiling to its AI music platform. The update marks the biggest feature expansion since Suno V5 Studio launched earlier this year.

For the broader landscape, see our complete producer guide to AI music and audio in 2026.

What Happened

Suno v5.5 introduces three features that shift the platform from generic generation toward personalized music creation. Voices lets Pro and Premier subscribers clone their own singing voice and use it directly in generated tracks. Custom Models allow users to upload their own songs as stylistic references, teaching the system a specific sonic direction. My Taste, available to all users, learns preferred genres and moods over time to adapt recommendations.

Why It Matters

Voice cloning changes the creator dynamic. Instead of choosing from preset voices, musicians can now generate tracks that sound like them. Suno includes a verification process that matches a live singing sample to a spoken phrase, and cloned voices remain private to the creator who made them. This addresses a core complaint from artists who felt AI music sounded generic and interchangeable.

Custom Models take personalization further. Pro and Premier users can create up to three custom models by grouping their own recordings as training references. The result is generated music that aligns with a specific artistic identity rather than a generic style tag. Combined with the Warner Music Group partnership announced in November 2025, this positions Suno for tighter integration with professional music workflows.

Key Details

  • Voices: Clone your singing voice for use in generated tracks. Pro and Premier only. Voice sharing planned for a future update.
  • Custom Models: Upload original songs to train up to 3 personalized style models. Pro and Premier only.
  • My Taste: Background system that learns your genre and mood preferences over time. Available to all users.
  • Base model: v5.5 is described as Suno's most expressive model to date, independent of the personalization features.

What to Do Next

Pro and Premier subscribers can access Voices and Custom Models now at suno.com. Start with voice cloning by recording a short singing sample through the Voices tab. For Custom Models, upload 3 to 5 of your own tracks that represent the style you want the system to learn. Free users get My Taste automatically. The timing is notable: Google Lyria 3 Pro launched the same week with structural music generation, making this the most competitive stretch in AI music creation to date.