Voice cloning is free. You only pay for the voiceover
Recording a voiceover for every video means a microphone, a quiet room, ten takes and editing. Voice cloning solves this once: upload 1–3 minutes of recording, the AI creates a digital copy of your voice, and from then on you just paste text and the voiceover sounds like you.
On NeuralBox, creating a clone costs no tokens at all, up to 10 voices per account. You pay only for the voiceover itself: from $0.20 per 1,000 characters. Under the hood is ElevenLabs, the world leader in speech synthesis.
How it works
Upload 1–3 minutes of clean voice recording, and the AI creates a digital copy in a couple of minutes. The voice then appears in the list next to the ready-made ElevenLabs voices: pick it, paste your text, and the audio is ready in seconds. Dozens of languages are supported, and the same voice can read texts in different languages.
If your own voice isn't enough or you need a character, there's Voice Design: creating a voice from a text description. Write “a deep narrator's voice for a fantasy saga, unhurried, slightly husky” and you get that voice from scratch.
What the voiceover costs
Cloning itself costs 0 tokens. Only speech synthesis is billed, by the number of characters in the text. Prices are on the Basic plan ($5 = 150,000 tokens); on higher plans tokens are up to 40% cheaper. One minute of speech is roughly 1,000 characters.
| Model | Price per 1,000 characters | Limit per request |
|---|---|---|
| Flash v2.5 | 6,000 tokens ≈ $0.20 | up to 40,000 characters — a whole book chapter |
| Multilingual v2 | 12,000 tokens ≈ $0.40 | up to 10,000 characters |
| Eleven v3 | 12,000 tokens ≈ $0.40 | up to 5,000 characters; the most expressive: emotions, whispers, laughter |
| Voice cloning | free | up to 10 voices per account (clones and Voice Design combined) |
What people clone their voice for
The most common case is voicing videos in your own voice without recording: clone your voice once, and from then on the voiceover for YouTube, Reels and tutorials is done by pasting text, with no microphone and no retakes.
The second is audiobooks: the Flash model takes up to 40,000 characters per request, so a chapter is voiced in one go, in one voice, at $0.20 per 1,000 characters. The third is a content pipeline: podcasts, courses and ads come out regularly and sound like one recognizable voice, even when the author has no time to record.
Ethics: whose voice you can clone
Clone only your own voice, or the voice of a person who has given explicit consent. Creating copies of other people's voices without permission is not allowed: it violates the terms of service and the law. You use the generated audio within the terms of service.
How to try it
Sign in with Telegram, Google or email, top up your balance (from $5 by card via Stripe, from $2 in crypto) and open the Text-to-Speech tab. Record or upload 1–3 minutes of clean speech — the clone is ready in a couple of minutes, for free.
A detailed look at the models, prices and an FAQ are on the Text-to-speech and voice cloning page.
Clone your voice for free
Creating a clone — 0 tokens, voiceover — from $0.20 per 1,000 characters. No subscription.
Create a voice