Upload a book you own
EPUB, PDF, or TXT. Novels, indie releases, translated fiction, web novel exports — anything with real text in it.
Full cast
Most text-to-speech apps read a whole novel in one flat voice. Lyraboo casts the book — narrator, hero, villain, everyone — and hands you a full-cast audiobook. Free to start on Android.
No voice-assignment screens, no per-character setup. Upload, wait a minute or two, press play.
EPUB, PDF, or TXT. Novels, indie releases, translated fiction, web novel exports — anything with real text in it.
It reads the whole text, separates narration from dialogue, works out who speaks each line, and assigns every character their own distinct voice.
Chapters, speed control, background playback. The first chapter is ready in about 1–2 minutes; the rest generates while you listen.
Standard text-to-speech treats a novel like a document: one synthetic narrator, start to finish. It works for a two-page memo. Across four hundred pages of fiction it fails in a specific way — dialogue stops being dialogue. Three characters argue in a room and every line arrives in the same register, so you lose track of who said what and start rewinding.
A full cast fixes that structurally rather than cosmetically. When the hero, the villain, and the narrator each hold their own voice, an argument sounds like an argument. You stop parsing the text and start following the story — which is the entire reason people listen to fiction instead of reading it.
Most multi-voice AI tools are built for the other side of the transaction: an author uploads a manuscript and exports a finished audiobook to sell. Useful, but that is a production tool with a render queue and an export button.
Lyraboo is built for the person who wants to listen. Your library lives in the app, audio generates chapter by chapter in the background while you play the earlier ones, and progress syncs the way an audiobook app should. You are not producing anything — you are listening to a book that was never going to get an audiobook edition.
Worth being straight about the edges. Voices are cast automatically and cannot be hand-picked. Attribution can slip in long stretches of unattributed dialogue. Scanned PDFs with no real text underneath have nothing to read — you need a text-based file. And narration is English only for now.
New to the idea? Start with the guide on how to turn a book into an audiobook, or go straight to EPUB and PDF conversion.
It reads the surrounding text the way you do. Dialogue tags (“said Marya”), the flow of an exchange, and who was addressed last are all used to attribute each line. It is not perfect — a long unattributed back-and-forth can drift — but across a full novel it holds up well enough that conversations stay followable.
No — casting is automatic. You upload the book and get a cast back. That is the whole point of the app: no setup screens, no assigning voices line by line before you can listen to anything.
Not really, and it does not force it. A book with no dialogue simply reads in the narrator’s voice. The full cast matters for fiction — novels, web serials, fanfiction, anything dialogue-heavy.
Close, but honest answer: no. A produced audio drama has human actors, sound design, and music. This is AI narration with a separate voice per character — much closer to a full-cast reading than to a single synthetic narrator, but it is not a studio production, and it costs a fraction of one.
Not yet — Lyraboo is currently Android-only. iOS is on the roadmap.
Free to start: every account gets free listening hours each month, no card required. Hours are spent once, when a chapter is first generated — replaying audio you already made is free forever.
Upload a book you own and let the app turn it into a multi-voice audiobook automatically.