Suno's New Speech Mode Narrates Over Original Music
Suno, the AI music app, opened a public beta of Speech, which turns a script or a one-line idea into a spoken recording with original background music.
Suno now makes spoken audio as well as songs. Speech, which opened to everyone as a public beta on October 1, gives you one track: an AI voice reading your words over background music written for it. Suno says its model generates the voice and the music together, in a single pass.
How Speech works
According to Suno's announcement, you type an idea or something you've written, then describe the voice and the musical style you want. You'll find it in Suno's web and mobile apps under the Create tab, then Speech. On mobile, Suno's post on X says to update the app first.
The Verge reports there are two ways in. Simple mode takes a plain description in a prompt box, like "a pirate captain rallying his crew," and builds the piece from that. Advanced mode lets you paste your own script when you already know exactly what it should say, and adds controls for the voice's gender, the speaking style, and how much each generation varies.
The music is on by default, and a toggle turns it off if you only want a plain voiceover. A single piece can run up to about eight minutes, according to both The Verge and Brief IA.
Suno's release notes pitch it with "bedtime stories over soft piano, hype speeches over stadium drums, ASMR grocery lists and more." Suno also posted a full walkthrough:
What Suno built it for
Jack Brody, Suno's chief product officer, is quoted on the launch. "Music will always be at the heart of Suno and what we build," he said. "At the same time, our vision has always extended to other forms of human expression."
The examples in the announcement are small and personal. While building Speech, the team "turned friends' texts into wildly overproduced dramatic readings," gave "ordinary voice notes unnecessarily epic scores," and made "meditations, poems, pep talks, and bedtime stories for our kids."
That lines up with how people already use the music side. Suno says that every day, people make songs for birthdays, weddings, inside jokes, faith and worship, their kids and their friends. Speech applies the same idea to the spoken word.
The Verge also points to a business reason: Suno's music generator has drawn a string of lawsuits, and branching into voice is likely a way to diversify.
"Beta really does mean beta"
Suno spent the past month testing Speech with a small group of users before opening it to everyone, and it calls this a true beta. "Occasionally, British accents can wander off to Australia and back," Brody said. "Dramatic pauses may be _very_ dramatic."
Suno says it will keep improving Speech based on what users report. Neither Suno's announcement nor its release notes mention pricing or credits for Speech.
Where it fits next to other voice tools
AI that reads text aloud has been around for a while. The Verge notes that DeepMind has worked on computer-generated speech for a decade, and that Adobe and ElevenLabs both offer text-to-speech tools.
Suno adds the background music, which its model writes in the same pass as the voice.
Voice tools like ElevenLabs already read scripts aloud. Suno also writes the music that plays under the voice, which turns a kid's bedtime story into a soft-piano audiobook in one step.
The use I find most convincing is the one Suno leans on: a short piece made for an audience of one, like a pep talk for a friend or a story for your kid. Suno songs already get made for exactly those moments, and a little roughness in the voice is easier to forgive there than in anything you'd put in front of a client.
Comments
Sign in to join the conversation.