AI music maker Suno now generates spoken words with its new Speech feature

Suno, the AI music platform, has launched a feature called Speech that generates spoken voiceovers from a script or a text description. It is in public beta on Suno's web and mobile apps, and it can produce a voice and background music as a single track.

Voice and music in one track

Suno describes Speech as "the first audio model that generates voice and music together as one cohesive track." Chief product officer Jack Brody said the company still treats music as its core: "Music will always be at the heart of Suno and what we build. At the same time, our vision has always extended to other forms of human expression."

Background music is optional: a toggle turns it off if you only want clean speech. Suno suggests the combination suits cases such as a calming soundtrack for poems or a more energetic backdrop for dramatic voiceovers and speeches.

How Speech works

Users open the Create tab and select the Speech option. Simple mode turns a prompt description, such as "a pirate captain rallying his crew," into audio; Advanced mode takes a custom script when the exact wording is already written. Advanced settings also adjust the gender of the voice, the speech style and how much variety each generation will have. A single generation can run for about eight minutes at most.

A crowded field and beta caveats

AI speech is hardly new: Google's DeepMind has experimented with deep-learning speech synthesis for a decade, Adobe has a text-to-speech tool, and ElevenLabs has become one of the most recognizable speech platforms since its 2023 launch. For Suno, the feature also looks like a way to diversify beyond music generation, which has attracted numerous lawsuits.

The company admits the feature is far from perfect. "Beta really does mean beta," said Brody, noting that "occasionally, British accents can wander off to Australia and back" and that dramatic pauses "may be very dramatic." Suno says it will keep improving the model around user feedback.