Voice cloningPaid plans
Voice cloning for audiobook authors: your book in your voice, no studio
Readers like hearing a book in its author's voice. Record a short sample of your reading, then let your cloned voice read the chapters from the text you paste in. For novelists, memoir and self-development writers and independent publishers who have neither weeks to spend in a studio nor the budget for one.
- Your voice, or written permission
- Private to your workspace
- Delete it at any time
Voice cloning is included on paid plans from Starter. Clone only your own voice or a voice you have written permission to use.
See the workflow
Building your voiceNow reading a new script:
كان البيت القديم يقف في آخر الزقاق، بابه أزرق ونوافذه عالية. في شتاء عام ألف وتسعمئة وأربعة وثمانين عدتُ إليه بعد غياب سبع سنوات. هل تغيّر شيء؟ لا أدري.
Hear it yourself
Scripts for recording and testing your narration sample
Read the first script aloud when recording, and the second only if your book has colloquial dialogue in the dialect you speak, then paste the third to hear your cloned voice read a book opening.
A calm narration sample
Modern Standard Arabicكان البيت القديم يقف في آخر الزقاق، بابه أزرق ونوافذه عالية. في شتاء عام ألف وتسعمئة وأربعة وثمانين عدتُ إليه بعد غياب سبع سنوات. هل تغيّر شيء؟ لا أدري. لكن رائحة القهوة كانت كما هي، والدرج ما زال يئنّ تحت قدميّ!
What it says: The old house stood at the end of the alley, its door blue and its windows high. In the winter of nineteen eighty-four I returned to it after seven years away. Had anything changed? I do not know. But the smell of coffee was just as it had been, and the stairs still groaned under my feet!
A colloquial dialogue sample from a novel
Egyptian Arabicبصّ يا عم حسن، أنا مش عايز أسمع الكلام ده تاني. إزاي تبيع الدكان من غير ما تقول لي؟ ده تعب أبويا وجدي! طيب، خلاص، اللي حصل حصل. دلوقتي نقعد ونشوف هنعمل إيه.
What it says: Look, Uncle Hassan, I don't want to hear that again. How could you sell the shop without telling me? That was my father's and grandfather's life's work! Fine, all right, what's done is done. Now let's sit down and see what we're going to do.
First test: the opening of the book
Modern Standard Arabicالفصل الأول. لم أكتب هذا الكتاب لأعلّم أحدًا شيئًا. كتبته لأنني وعدت أمي، قبل عشرين عامًا، أن أحكي ما جرى كما جرى. وها أنا أفي بالوعد، متأخرًا كعادتي.
What it says: Chapter one. I did not write this book to teach anyone anything. I wrote it because I promised my mother, twenty years ago, that I would tell what happened as it happened. And here I am keeping the promise, late as usual.
Every recording is real Vocla output in a public library voice, published unedited.
The problem
A three-hundred-page book means weeks at the microphone
One clean finished hour takes hours of reading, retakes and editing, and a single book runs to many hours.
A writer who is not a trained reader tires after an hour, so chapter ten ends up in a different tone from chapter one.
Then comes proofing: a character's name said two ways, a word read with the wrong case ending, a sentence dropped. Going back to the studio for ten sentences costs a full day, and the new take does not match the old one.
A professional narrator is an excellent option, but it is not your voice. In a memoir or a book about personal experience, listeners expect to hear the person the story happened to.
Why Vocla
Why a cloned voice suits a whole book
Your voice from the first chapter to the last
The voice is built from your sample, so it does not tire and its pitch does not shift between sessions. The final chapter comes out in the tone the book began with.
One-click tashkeel with every change reviewed
Add tashkeel to a paragraph or an ambiguous word and review what changed, or listen to the possible readings of a word and pick the right one. Tashkeel marks are never deducted from your credits.
A pronunciation dictionary for characters and places
Add the hero's name, the village and the foreign term once, and each is read the same way in every chapter.
Fix a sentence, not a chapter
Change the sentence, regenerate its paragraph with the same settings, then swap it into the chapter file. Everything you generate is kept in your history.
Up to 10,000 characters per generation
That is about ten minutes of audio, a good size for a scene or a section of a chapter. A long chapter is generated in parts that you then join.
How it works
From manuscript to narrated chapters
Step 1
Record the sample in your narration tone
Read passages from the book itself, at the pace you want for the whole thing, in a quiet room. One to two minutes across a few clips is enough, and ten seconds is the minimum.
Step 2
Prepare the text for the ear
Vocla does not import Word or PDF files, so copy the text and paste it in. Remove footnotes and page numbers, write numbers and dates out as words, and turn tables into sentences.
Step 3
Fill the pronunciation dictionary before chapter one
Collect proper names and foreign words from the entire manuscript and settle how each is said, so you do not discover the inconsistency in chapter seven.
Step 4
Generate chapter by chapter and listen
Split each chapter into parts of no more than 10,000 characters at the end of a scene or paragraph. Listen to every part in full while following the text, and correct whatever stumbles.
Step 5
Download WAV and name the files
Download the parts as WAV and number them by chapter and part, then assemble them in an audio editor and add the silences, the opening and the closing credits.
Listen to the voices
Hear Vocla's voices before you add your own
These are voices from Vocla's public library, not cloned voices. Your cloned voice stays private to your workspace and reads any script you write, the way these voices do.
- SafaaArabicModern StandardFemale · Narrative
- ReemArabicModern StandardFemale · Family
- MazenArabicModern StandardMale · Narrative
- AmaniArabicModern StandardFemale · Narrative
- AdelArabicModern StandardMale · Narrative
- LaylaArabicModern StandardFemale · Narrative
A voice belongs to a person. Cloning starts with their permission.
Voice cloning is powerful, so it runs under clear rules. Every cloned voice on Vocla must be your own voice or one you have explicit, written permission to use, and we act on every report of misuse.
Read the voice consent policyText to speech in the matching dialects
Cost
What does an audiobook in your voice cost?
Cost follows the character count of the manuscript, not its page count, and a thousand characters is about one minute of audio. Get the character count from your word processor to estimate the book before you begin, and remember that tashkeel marks are not counted and that a sentence you regenerate is charged for its own characters only.
Voice cloning, commercial usage rights and WAV downloads all start with the Starter plan at $4.99 a month with 5,000 credits. A full book usually needs the Professional plan's 80,000 credits a month, or its chapters spread over more than one month, and you can buy a credit top-up when you need one.
Compare plansThe numbers
- Voice cloning is included from Starter, at $4.99 a month. It is not part of the Free plan.
- Cloned voices you can keep: 1 on Starter, 5 on Professional and 15 on Business.
- Audio read by your cloned voice is billed like any text to speech: 800 to 1,200 credits per 1,000 characters.
Figures use the current plans and default credit rates. The exact cost is always shown in the app before you confirm.
Publishing this for a business?
The Free plan is for personal use and evaluation. Audio and video you publish for a brand, a client or a monetised channel need commercial usage rights, which start with the Starter plan.
Questions people ask
Can I record my audiobook in my own voice without a studio?
Yes. You record a sample of one or two minutes in a quiet room, and Vocla builds a cloned voice that reads your chapters from the text. You still need time to listen and correct, but far less than reading the whole book at a microphone and redoing every stumble.
Do audiobook platforms accept AI narration?
Rules differ between platforms and they change; some ask you to disclose generated narration or set conditions for it. We cannot guarantee that any distributor will accept it, so read your distributor's terms before generating the whole book, and ask in writing if they are unclear.
How many characters can it read in one go?
Up to 10,000 characters per generation, about ten minutes of audio. Split a long chapter at the ends of scenes or paragraphs, generate each part, then join the parts in an audio editor. Splitting also makes it easy to replace a single part when you correct something.
How do I control how character and place names are pronounced?
Two ways: add tashkeel to the name in the text, or add it to the pronunciation dictionary with the reading you want, and it applies wherever the name appears. Write foreign names in Arabic letters as they sound. You can also hear the possible readings of a word and pick one before generating.
Does the cloned voice act out different characters?
It does not change its voice per character the way an actor does. It reads in your voice, with an emotional range that follows the sample and your expressiveness setting. That suits single-narrator books; multi-voice productions have a different tool, text to dialogue with library voices.
Can I sell an audiobook generated in my cloned voice?
Commercial usage rights for what you generate are included from the Starter plan up, provided the voice is your own or one whose owner gave written permission. Two things outside Vocla remain yours to check: that you hold the audio rights to the text, and that your distributor accepts generated narration.
Go deeper
- The tool: Voice cloningClone your own voice from a short recording
- Plans and pricesWhat each plan includes, side by side
- Voice libraryPlay real samples by dialect
- How to Make an Arabic Audiobook with AI Narration: From Manuscript to Chapter FilesRead the guide
- Voice Cloning Sample Tips: How to Record a Sample That Sounds Like YouRead the guide
- How to Clone Your Own Voice in Arabic, Step by StepRead the guide
Related solutions
Other pages for a neighbouring job.
Arabic text to speech for audiobooks, chapter by chapter
Turn your manuscript into an audiobook chapter by chapter, in one narrator's voice.Arabic text to speech for kids' stories and bedtime tales
Stories told by a warm narrator, in simple Modern Standard Arabic or the dialect of home.Voice cloning for content creators: YouTube, podcasts and courses in your voice
Your own voice on every video and episode, from a script you type, not a recording session.AI brand voice for Arabic content: clone a voice or design one
One voice your customers recognise in ads, IVR, app prompts and social clips.Arabic AI voice-over generator for any video
An Arabic voice-over for any video from a written script, with no voice artist or studio.Arabic male AI voice generator: Modern Standard Arabic and dialects
Male voices in Modern Standard Arabic and the dialects, from measured to youthful.
Start in minutes
Give your content a voice today
Start free with 500͏ credits every month. No card required.
- 500͏ free credits every month
- No card required
- Arabic and English interface