THE NEWSPAPER / 01A new chapter for AI Music Index

The Music Index

News / The reading room

AI tools

Suno starts talking — Speech creates voice and music in one take

Suno Speech generates spoken voice and original music together. Write a poem, a bedtime story or a pep talk, describe the voice and music — and give even a shopping list an absurdly epic soundtrack.

A microphone with voice waves becoming music and film strips.
A microphone with voice waves becoming music and film strips. Editorial AI illustration in a newspaper print style.

The new beta is available on the web and mobile. It is designed for stories, meditations, poems and dramatic readings rather than conventional songs — although the voice can still develop an unexpected accent and a taste for very dramatic pauses.

What happened?

Suno launched Speech in beta on 1 October. You write an idea, a poem or your own text, then describe the voice and musical style you want. The system generates speech and background music together in the same audio track.

Suno calls it its first model to generate voice and music as a whole. The feature was tested with a smaller group and is now opening more widely on the web and in the mobile app.

What can you do with it?

The obvious uses include bedtime stories, meditations, poems, presentations and short stories. The more playful side is adding trailer drums to an ordinary voice message, or reading the weekly shopping list as if humanity’s future depends on two avocados and some oat milk.

This goes beyond music generation. Suno is moving towards personal audio content in which narration, pacing and the soundtrack are shaped together.

Beta can mean unexpected results

Suno itself warns that accents may wander between countries and that dramatic pauses can become very dramatic. Listen carefully to the result before publishing it.

The launch announcement does not describe every detail of the voices’ training material, commercial rights or the specific safeguards against imitating real people. Check the current terms before using the material beyond a private experiment.

What does this mean for music makers?

Speech could be a quick sketching tool for spoken word, intros, interludes and the stories around a song. Do not export blindly: treat the result as a draft that needs human review.

  • Write your own text or use text you have permission to use.
  • Avoid instructions intended to imitate an identifiable real person.
  • Listen for mispronunciations, odd pauses and music that clashes with the words.
  • Keep the prompt, model, date and the version you actually use.

A closer look

When speech and music are generated together, the system can coordinate emphasis, pauses and musical movement. This differs from creating a voice file first and then placing a separate piece of music underneath it in a DAW.

The trade-off is less control. If one sentence or chord is wrong, the entire track may need to be regenerated or split up for further editing. Separate voice and music tracks remain easier to repair when you want to control every detail.

GO TO THE SOURCE

Sources & background

Sources from the existing article. This reading edition preserves the original reporting date.

  1. Suno: Introducing Speech (beta)about.suno.com
  2. Suno release notes: Speech on web and mobileabout.suno.com
↑ Back to the beginning