Winston
English · Narration
Give the narrator and every speaking character in your book a distinct voice. FullCastBook works out who says each line; you approve the cast, direct the delivery and export.
my-story.txt
Text
Reading your file…
Lewis Carroll
Alice's Adventures in Wonderland
Alice asks the grinning Cheshire Cat for directions — and gets none.
The Cat only grinned when it saw Alice. It looked good-natured, she thought; still it had very long claws and a great many teeth, so she felt it ought to be treated with respect.
Step-by-step guide
Free to start: no card, nothing to install.
Multi-Voice Audiobook
What you get, from manuscript to finished audiobook.
Use cases
Where authors and storytellers put it to work.



Side by side
What changes when the cast is part of the tool instead of an afterthought.
| FullCastBook | Single-voice TTS | |
|---|---|---|
| A distinct voice per character | Yes | No |
| Speakers detected from the manuscript | Yes | No |
| Chapters from the contents page | Yes | Varies by tool |
| Direction notes and sound tags | In brackets, in the text | Varies by tool |
| Regenerate a single line | Yes | Varies by tool |
| M4B with chapter markers | Yes | Varies by tool |
| Free to start | Free credits, no card | Varies by tool |
Demo performances. In the studio you cast from standard and premium voice lists.
English · Narration
English · Narration
English · Narration
English · Character
English · Narration
English · Narration
English · Narration
English · Narration
English · Narration
English · Character
In depth
A multi-voice audiobook gives the narrator and each speaking character their own voice, so a conversation sounds like a cast instead of one performer changing pitch between lines. Listeners always know who is talking, even in a long scene with four people in the room.
FullCastBook is a multi-voice audiobook generator that builds that cast from your manuscript. It finds the chapters, works out who speaks every line of dialogue, and hands you a cast list to fill with voices. You choose the voices, add direction where a line needs it, and export a finished audiobook with different voices for each character.
When people look for a multiple narrators audiobook, they usually mean one of three production styles. A single narrator reads everything and suggests characters with small shifts in tone. A duet splits the book between two narrators, often one per point-of-view character. A full cast gives every speaking part its own performer, closer to a radio drama.
With multiple narrators, each extra voice adds booking, direction and editing time, which is why full casts are rare outside big releases. In an AI audiobook with multiple voices, the extra parts are a casting decision, not a budget line, so you can choose the style that suits the book. Many authors land in the middle: one steady narrator for the prose and character voices only inside quotation marks. For a closer look at the trade-offs, see full cast vs single narrator.
Before you cast anyone, FullCastBook reads the manuscript the way an editor would. Using dialogue tags such as "she said" and the context around each line, it assigns every line to the narrator or to a named character.
Detection does most of the work, but it is a first pass, not a final edit. Unattributed back-and-forth, characters who share a name, or a narrator who quotes someone in passing can be read the wrong way. Skim the dialogue-heavy chapters, change the speaker on any line that's off, and the rest of the multi-voice audiobook follows the cast you set.
The easiest cast to listen to has clear contrast. Give the main characters voices that differ in age, pitch or accent, keep the narrator neutral and easy to listen to for hours, and save the most striking voices for smaller parts that only appear now and then.
Listen to each voice before you commit, and listen to a full chapter rather than a single line: a voice that sounds great in one sentence can tire the ear over twenty minutes. If a voice stops fitting, swap it in the cast list and regenerate. Once a character sounds right, keep that voice for every book in the series. For a feel of what different voices can do, hear the example performances on the home page; in the studio you cast from the standard and premium voice lists. For the craft of making characters sound distinct, see how to do character voices.
Voices read the text as written, so the text is where you direct them. Put a note in brackets before a line, like [whisper, like sharing a secret] or [angry, through clenched teeth], and the delivery follows. Sound tags such as [laugh], [sigh] or [breathe] add the small human noises that make dialogue feel alive.
Use direction sparingly. Most lines read well on their own; a handful of notes at the turning points of a scene will do more than marking every sentence. If a line still lands wrong, rewrite it slightly or move the note, then generate it again.
When the chapters sound right, export an M4B with chapter markers for audiobook players, one MP3 for the whole book, or a ZIP with a file per chapter. The files are plain audio with no DRM added, and they are yours to use. For a short scene or script rather than a whole book, try multi-voice text to speech.
Each store has its own policy on AI-narrated audio, so check it before you upload. Audible and ACX don't currently accept AI narration. Want to know the running time and credits first? Try the audiobook cost calculator, see the plans, or open the AI audiobook generator.
Usually a full-cast audiobook, or a dramatized one when it adds sound design; a book shared by two narrators is often called a duet. A multi-voice audiobook made with FullCastBook is the full-cast kind: a narrator plus a voice for each character.
It reads the manuscript, splits it into chapters and attributes every line of dialogue to the narrator or to a named character, using dialogue tags and the flow of each conversation. You review the result before anything is generated.
Yes. Change the speaker on that line in the studio before you generate. Detection is a strong first pass, so it's worth skimming chapters with long unattributed conversations.
Yes. The multi-voice audiobook generator puts every speaking character it detects on the cast list. Give the main ones their own voices and let walk-on parts share a few supporting voices, which keeps your most distinctive voices for the roles that matter.
Yes. Swap the voice in the cast list, then regenerate the chapters you want to update. Regenerating voices that text again, so it uses credits again.
Yes. Keep the same narrator and character voices for every book in a series, so a multi-voice audiobook series sounds like one production.
Audible and ACX don't currently accept AI-narrated audiobooks. The files you export are yours, so you can publish them wherever you hold the rights to the text and the store accepts AI narration. Check each store's policy first.
Upload your manuscript, cast the voices and export every chapter of your multi-voice audiobook. Free to start.