How to voice text with your own voice in Text to Speech
In Text to Speech, pick your voice on the My voices tab and tap Generate. A MiniMax voice costs 12 tokens per 1,000 characters, a Qwen 3 voice about 11.
Updated November 2, 20263 min read
First you need a ready voice in My Voices. If you don't have one yet, create it from a recording of 10 seconds to 5 minutes: how to clone your voice. A voice made in the welcome tour is already there.
How to voice a text in your voice
On the Tools tab, open Text to Speech.
On the Text tab, enter up to 10,000 characters and tap Continue.
On the Voice tab, open the My voices tab and tap your voice. The badge next to it shows the provider, MiniMax or Qwen 3, and the line below shows the language.
For a MiniMax voice, set the emotion, speed and pitch below the list if needed. Tap Continue.
On the Result tab, check the estimated cost and tap Generate.
The play button next to your voice plays the original recording. To hear the clone itself, voice a short phrase first: it costs about 1 token.
What you can set up
| Your voice | Settings | Tags in the text |
|---|---|---|
| MiniMax | emotion, speed from 0.5 to 2, pitch from −12 to +12 | (laughs), (sighs) and others, a pause <#1#> |
| Qwen 3 | none | none |
Text instructions work only with OpenAI voices, so you can't use them with your own voice.
In which languages your voice speaks
A Qwen 3 voice speaks 10 languages: English, Russian, Chinese, German, French, Spanish, Italian, Japanese, Korean and Portuguese. A MiniMax voice sounds most natural in the language the sample was recorded in. In another language it keeps the accent of the original.
How much it costs
| Your voice | Price | One minute of speech |
|---|---|---|
| MiniMax | 12 tokens per 1,000 characters | about 11 tokens |
| Qwen 3 | about 11 tokens per 1,000 characters | about 10 tokens |
Creating the voice is paid separately and only once: 180 tokens for MiniMax, about 1 token for Qwen 3. Some MiniMax voices created long ago were never activated. For such a voice, the Result tab shows "Includes 180 tokens for first-time voice activation, one-time only", and the first generation costs 180 tokens more. After that the voice costs the usual price.
Where else your voice works
Translate audio: in the settings, open Voiceover, choose MiniMax or Qwen 3, then My MiniMax voice or My Qwen voice. The translation is read in your voice. More: Translate audio settings.
AI agents connected over MCP, such as Claude and ChatGPT, see the same voices and can voice text with them.
Voice Changer, AI Calls and Translate Video don't use saved voices.
Frequently asked questions
My voice isn't on the My voices tab. Why?
If it's greyed out with the Cloning... badge, wait until cloning is done. With the Failed badge it can't be used: see voice cloning failed. If the voice is missing completely, check that you are signed in to the same account where you created it.
Can I use my voice on my computer if I created it on my phone?
Yes. Voices are stored in the account. Sign in to the same account on web.vocalingo.app, and the voice is in My voices. Sign in with Apple is available only on iOS, so on the web use the email or Google sign-in of that account.
Contact support
Describe the problem and we will reply by email. Add the file name or job ID if you have one.