Suno Opens Speech Beta to Generate Voice With Music

Suno is best known for turning text prompts into songs. On October 1 the company opened a public beta of Speech, an audio model that generates spoken voice and original background music together in a single track. The feature is live on Suno's web and mobile apps for all users, after a month of testing with a small group.

"Music will always be at the heart of Suno and what we build," Suno chief product officer Jack Brody wrote in the announcement. "Today, we're expanding what's possible in Suno with Speech: the first audio model that generates voice and music together as one cohesive track."

What Suno shipped

Speech works from a prompt or a full script. A Simple mode takes a description such as "a pirate captain rallying his crew" and generates the performance; an Advanced mode accepts a custom script and lets users pick voice gender, speech style, and how much variation each generation produces. Background music is optional and can be switched off with a toggle for clean narration. Suno caps a single Speech generation at roughly eight minutes.

The company is explicit that the model is early. Brody warned that "British accents can wander off to Australia and back," and that dramatic pauses may land harder than intended. Speech sits in beta alongside a music generator that has drawn repeated copyright lawsuits, and a broader voice offering is an obvious diversification play.

Voice AI gets another heavyweight

AI speech synthesis is a crowded category. DeepMind has researched speech synthesis for over a decade, Adobe ships a text-to-speech tool, and ElevenLabs has become the best-known dedicated platform since 2023, as The Verge notes. What separates Speech is the packaging: instead of a plain voice track, it returns a finished piece with a score attached. That is a consumer-creator wedge rather than a developer API.

The distinction matters to builders. Suno is not positioning Speech as a model endpoint for agent pipelines; it is positioning it as a creation surface inside a subscription app. Teams that need programmatic voice for agents, IVR, or localization still reach for dedicated text-to-speech providers with predictable latency, voice controls, and per-character pricing.

What it means for builders

• Audio generation is converging. Voice and music are becoming one output rather than two separate services.
• Consumer-first launches leave the API layer open. Suno's Speech beta creates demand for generated voice without satisfying programmatic use cases.
• Beta quality still gates production use. Accent drift and pacing variance are fine for creative work and unacceptable for customer-facing agents.
• Watch the licensing model. Music-adjacent generation carries rights exposure that plain narration does not.

Suno has not announced a standalone price for Speech or an API for it. For now the beta is available inside Suno's existing web and mobile apps.