ElevenLabs turns any ePub into a cast-narrated audiobook

Turning a written manuscript into an audiobook used to mean booking studio time, auditioning voice actors, and burning through thousands of dollars over several months.

ElevenLabs wants to shrink that entire ordeal into a lunch break. The voice AI company has rolled out a full-suite audiobook creation workflow designed to generate studio-grade audiobooks directly from digital manuscripts.

Automatic casting and instant directing

The marquee feature here is character casting. Instead of forcing a single synthetic voice to drone through an entire novel, the system scans your book and automatically identifies distinct characters.

It then assigns matched voice profiles to each character and previews how they sound using actual lines of dialogue from the manuscript. If a villain sounds too gentle or a protagonist sounds flat, you can swap voices on the fly until the cast fits your vision.

Creators get access to over 10,000 voices, alongside the ability to clone their own voice in seconds. The studio editor also lets authors fine-tune tricky pronunciations, layer in sound effects, and drop in background music.

The tool supports more than 90 languages and takes standard ePub uploads. Once the audio is baked, authors can export the files anywhere or publish directly to the ElevenReader ecosystem within minutes to monetize their work.

Why it matters

For indie authors and smaller publishers, this essentially removes the financial barrier to the audiobook market.

Millions of backlist titles never get audio adaptations simply because traditional production costs make it impossible to break even. Dropping production time to minutes and costs near zero unlocks an enormous wave of long-tail spoken content.

We think ElevenLabs is making a very sharp vertical move here. By handling everything from manuscript parsing and dynamic casting to sound effects and distribution, they are not just selling an API—they are building the full publishing stack.

The real test will be long-form endurance. A voice can sound convincing in a thirty-second dialogue preview, but keeping listeners engaged across a twelve-hour epic without emotional fatigue is a different beast entirely. Still, for budget-conscious creators, this is an immediate no-brainer.