The announcement came in a company blog post written by chief product officer Jack Brody. Speech is built directly into the Suno product: users type in an idea, a poem, or something they have written, describe the voice and musical style they want, and the model produces spoken audio set to an original background score.
Two modes, up to eight minutes of speech
Speech ships with two ways to create. A Simple mode lets users describe what they want through a prompt box — Suno suggests examples like “a pirate captain rallying his crew” — while an Advanced mode accepts a custom script for those who know exactly what they want said, with granular controls over the voice’s gender, speech style, and vocal variety. Generated speech runs up to about eight minutes.
The musical layer is optional. A toggle switches the background score on or off, so the same tool doubles as a plain text-to-speech generator when music would get in the way. To try it, users select the Create tab and navigate to the Speech option on web or mobile.
A crowded field, a deliberate differentiator
AI-generated speech is hardly a new frontier. DeepMind has been experimenting with deep-learning speech synthesis for a decade, Adobe offers its own text-to-speech tools, and ElevenLabs has built a sizable business on voice generation since 2023. Suno’s pitch is integration: pairing voice with an automatically generated soundtrack in a single pass, rather than forcing creators to stitch two tools together.
The company is framing the launch around what it calls “creative entertainment,” a category it says will define the next wave of consumer technology. Its suggested use cases include bedtime stories over soft piano, hype speeches over stadium drums, pep talks, poems, and meditations — the same personal-creation culture, Brody wrote, that already drives users to make songs for birthdays, weddings, and inside jokes.
Beta quirks and a lawsuit backdrop
Suno is upfront that the feature is unfinished. “Beta really does mean beta,” Brody said, noting that British accents can occasionally drift into Australian ones and that “dramatic pauses may be very dramatic.” The company says it will refine Speech based on user feedback and expects the community to discover uses its own team never imagined.
The launch also lands at a strategic moment. Suno’s core music product remains tangled in copyright lawsuits from major record labels over how its AI was trained, making a second product line a useful hedge. Notably, the company’s September v6 music-model release was developed alongside Warner Music Group, BMG, and Believe — a sign Suno is simultaneously trying to build licensed relationships with the industry it has been fighting in court.
Whether Speech becomes a serious narration tool or stays a creative toy will depend on how quickly the beta’s quirks get ironed out. For now, it marks a clear expansion of Suno’s ambitions: from songs to the full soundtrack of human expression.
By Hamza Faisal
