Performed vs NarratedSome books want a reader. Some want a performer.

Our newest voices don't just read your sentences. They read the scene they sit in.

Performed is TomeVox's newest range of AI audiobook voices: 12 each in English (US and UK), Spanish and Swedish, that read the scene around each line and shape their tone, pacing and breath to it. When an English book slips into another language, French, Spanish, German, Japanese or almost any other, they pronounce those words correctly while keeping the narrator’s own accent. Narrated, our clear and natural range, has 50+ voices in 13 languages.

Take the line “I’m fine.” Said at a graveside, it is a lie told quietly. Said on a doorstep after a first date, it is almost a laugh. The words are identical; the reading is not, and only the paragraphs around the line tell you which reading is right. A listener hears that difference in the first second. Most text-to-speech reads each sentence on its own, so it gets the words right and the moment wrong: an even, polite voice delivering a confession and a grocery list in the same key.

Over a ten-hour novel, that flatness is what makes listeners drift off. It is also why so much AI narration still sounds like AI narration: it pronounces everything correctly and still does not seem to know what is happening in the story. Why AI audiobooks sound robotic goes through all four causes and the fix for each.

What is a performed audiobook reading?

A performed reading is one where the voice reacts to the story, not just to the sentence. Before it speaks a line, the voice takes in the passage around it: who is talking, what just happened, whether the scene is tense, tender or funny. It then sets its tone, its pace and even where it breathes to match. A threat slows down and drops. An argument speeds up and overlaps its own clauses. A joke gets the half-beat of timing before the punchline.

That is the difference between our two voice ranges. Narrated reads clearly and naturally, with a choice of a Classic or an Expressive style for the whole book. Performed, our newest generation of voices, decides how to deliver each line from the scene itself, so no style setting is needed. You don't need to mark up your manuscript; the voice works from the words of your book, exactly as you wrote them.

The quickest way to understand it is to listen. Below are two of our stock voices, one from each range:

What happens when an English book switches language?

A Performed voice pronounces foreign words correctly, in its own accent, and carries on in English without a seam. Plenty of English-language fiction does not stay in English: a grandmother who speaks French at the dinner table, a soldier’s German, a Spanish street name, a greeting in Japanese, a line of Latin on a gravestone. Voices that only know English tend to flatten those moments into English sounds, and the listener hears the author’s careful detail as a mistake.

In a passage like this, a Performed voice pronounces the French correctly, keeps the old woman’s warmth in it, and carries straight on into “the old woman said” without changing who is narrating. It does not put on a French accent; it keeps its own, the way a well-read English narrator would. The same holds for a phrase dropped mid-sentence, such as a card that reads Ich vermisse dich: the German is pronounced correctly, and the sentence keeps its rhythm around it. French and German are only examples: the same goes for Spanish, Italian, Portuguese, Russian, Japanese, Arabic and almost any other language your characters speak. For authors writing immigrant stories, historical fiction, thrillers set abroad or any book with a bilingual cast, this is the feature that tends to decide it.

If your whole book is in Spanish or Swedish, Performed has 12 voices for each. For any other language, Narrated covers 13 languages today, and our guide to foreign-language audiobooks covers translating an English book.

How is Performed different from Narrated?

Performed and Narrated are two separate voice ranges with the same files and the same service around them. Every book gets one voice, from one of the two ranges:

Performed gets a third revision round because a performance has more to tune: if a scene lands too hot or too cool for your taste, you tell us and we redo it. More languages, multi-voice casts and your own voice are coming to Performed; for now, if you want a full cast or your own voice, Narrated is the range that has them.

How much does a Performed audiobook cost?

Performed costs $149, $229 or $299 per book; Narrated costs $49, $79 or $99. Both are flat prices set by word count, so you know the price before you upload.²

Per hour of finished audio, Performed works out at roughly $21 for a 100,000-word novel, which runs to about 10.8 finished hours at around 9,300 words an hour; Narrated comes to about $7.¹ For what a human narrator or a DIY tool costs for the same book, see how much an audiobook costs.

Two comparisons authors ask us about. If you would otherwise make the book yourself in ElevenLabs’ self-serve app, we do the preparation, the regenerations and the checking for you, and on a full book it costs less. If you were looking at ElevenLabs Productions, their done-for-you service, Performed is the direct alternative: the same quality, if not better, for a fraction of the price and with a much faster turnaround.

Which should you pick for your book?

Pick Performed if your book is in English, Spanish or Swedish and lives on its scenes: dialogue-heavy fiction, romance, thrillers, literary fiction, memoir with emotional range, or any English book that slips into another language along the way.

Pick Narrated if your book is in one of our other 10 languages, if you want it read in your own voice, if it is straightforward nonfiction where clarity matters more than drama, or if the lowest price per book is the priority.

If you are unsure, you don’t have to guess. Your first chapter is free in whichever voice you pick, so you can upload once, listen to your own opening in a Performed voice, and decide. Our guide to choosing an audiobook voice covers the rest of that decision.

Your book already knows how it should sound. Performed listens.

Common questions about Performed

What is the TomeVox Performed edition?

Performed is TomeVox’s newest range of AI audiobook voices: 12 each in English (with US and UK accents), Spanish and Swedish. Each voice reads the scene around a line and shapes its tone, pacing and breath to it, and pronounces words from almost any other language inside an English book correctly, in its own accent.

How much does a Performed audiobook cost?

Performed costs $149 up to 60,000 words, $229 up to 100,000 words and $299 up to 150,000 words, plus $0.0015 per extra word beyond that. Narrated costs $49, $79 and $99 for the same lengths, plus $0.0005 per extra word over 150,000.

Can a Performed voice read phrases in other languages inside an English book?

Yes, in almost any language. A Performed voice reads a line of French dialogue, a Spanish street name, a German phrase or a Japanese greeting with the correct pronunciation, keeping its own accent, and carries on in English without breaking the sentence.

Which languages does Performed support?

Performed books can be in English (with US and UK accents), Spanish or Swedish. More languages, multi-voice casts and your own voice for Performed are coming soon. Narrated covers 13 languages today and offers your own voice.

Can I hear a Performed audiobook before I pay?

Yes. Upload your manuscript, pick a Performed voice, and your first chapter is free, with no card. You pay only if you buy the full audiobook, which is ready in about 24 hours.