AI Voice Clone: Make Your Own Voice Read Any Text
Updated
Short answer: VdoBloom AI Voice Clone takes a 3–15 second recording of your own voice and reads any text up to 2,000 characters in that voice. It speaks 10 languages (English, Chinese, Japanese, Korean, German, French, Russian, Portuguese, Spanish and Italian). A run costs 0.0012 credits per character (1 credit minimum) plus 6 credits to transcribe the sample, or nothing extra if you type what the sample says. You must confirm the voice is yours or that you have the speaker's permission.
Record straight in the browser or upload an MP3, WAV, M4A or similar file. Type the text you want spoken, pick a language and speed, tick the consent box and press Clone & Speak. The result plays in the page, downloads as an MP3 and is saved to My Creations.
It is built for creators who want their own voice on narration, tutorials, social clips and drafts without recording every line. Cloning celebrities, politicians or any other public figure, or anyone who has not agreed, is not allowed.
How it works
- 1
Record or upload your voice
Record 3–15 seconds in the browser (it stops by itself at 15 seconds) or upload a clip of that length. Use one speaker, clear speech and a quiet room.
- 2
Optionally type what you said
Typing the exact words of the sample skips automatic transcription and its 6-credit charge. Leave it empty and the sample is transcribed for you.
- 3
Write the text and pick a language
Paste up to 2,000 characters, choose one of the 10 languages or Auto, and set the speed from 0.5× to 2×.
- 4
Confirm consent and generate
Tick "This is my own voice, or I have the speaker's permission", check the price shown above the button and press Clone & Speak.
Quick facts
| AI Voice Clone | |
|---|---|
| Voice sample | 3 to 15 seconds, recorded in the browser or uploaded (max 25 MB) |
| Text per run | Up to 2,000 characters |
| Languages | English, Chinese, Japanese, Korean, German, French, Russian, Portuguese, Spanish, Italian, plus Auto |
| Not supported | Hindi, Tamil and other languages outside the list above |
| Price | 0.0012 credits per character (1 minimum) + 6 credits for sample transcription, or + 0 if you type the sample's words |
| Speed | 0.5× to 2× |
| Output | MP3, saved to My Creations |
| Consent | Required on every run and stored with the job |
Consent and whose voice you can clone
Only clone your own voice, or the voice of someone who has given you permission. Every run requires the checkbox "This is my own voice, or I have the speaker's permission", and that confirmation is stored with the job. Cloning celebrities, politicians or other public figures, or anyone without their agreement, is against the Terms of Service, and texts go through content moderation before anything is charged.
Read more: Terms of Service
How much does it cost?
The price is shown before you generate and is the same number that is charged. Examples:
| Run | Credits |
|---|---|
| Short text (up to 833 characters), sample words typed | 1 |
| Short text (up to 833 characters), sample auto-transcribed | 7 |
| Full 2,000-character text, sample auto-transcribed | 9 |
Read more: Credit packs and plans
Tips for a closer match
- Record 8–15 seconds of natural speech rather than 3 seconds of a single word.
- Avoid music, echo, other voices and background noise in the sample.
- Speak in the tone you want the result to have: calm samples give calm results.
- If the words of your sample are unusual (names, jargon), type them in yourself.
Voice Clone or Voice Design?
Voice Clone copies the sound of a recording you provide. Voice Design invents a new voice from a written description and needs no recording at all. Use Voice Design when you want a character or narrator voice that is not a real person.
Read more: Open Voice Design
Frequently asked questions
How long does my voice sample need to be?
Between 3 and 15 seconds. Longer, clean samples of natural speech usually give a closer match.
Can I clone a celebrity or a public figure?
No. You may only clone your own voice or a voice whose speaker has given you permission. Cloning celebrities, politicians or other public figures is against the Terms of Service.
Does it support Hindi or Tamil?
Not yet. The clone speaks English, Chinese, Japanese, Korean, German, French, Russian, Portuguese, Spanish and Italian. For Hindi or Tamil voiceovers, Text to Speech has stock voices in those languages.
Why is there a 6-credit transcription charge?
The model needs the exact words spoken in your sample. If you leave that box empty we transcribe the sample automatically, which costs 6 credits. Type the words yourself and the charge is 0.
What happens if a run fails?
The job is marked failed and its credits are refunded automatically.
Can I record in the browser?
Yes. Press Record, speak, and recording stops by itself after 15 seconds. You can also upload a file.
Where is my result saved?
It plays and downloads as an MP3 on the page and is saved to My Creations.