Voice Studio at /voice-studio makes a personal voice from a recording of 10 seconds to 3 minutes, then turns any text into an MP3 in that voice. Up to 5 voices, private to your account.
Voice Studio gives your business your own voice without a studio. You record yourself once, and from then on you write a text and get a recording in your voice: a greeting for the clinic phone line, narration for a short video, a voice message to customers. Open it from the account menu, Voice Studio, or go to /voice-studio.
1. Click Create a voice: Under My voices, click Create a voice and give it a Voice name, for example "my narration voice".
2. Add a sample: Upload an audio file (Choose an audio file, up to 50 MB) or click Record with microphone. The voice is made from 10 seconds to 3 minutes of speech; 30 to 60 seconds of one speaker, without music or noise, works best. From a longer recording you choose the part to use. Need something to read? gives you a text.
3. Confirm and create: Tick the box that this is your voice, or that you have explicit permission from its owner, and click Create my voice.
1. Write the text: Choose a ready voice and write under What would you like to say?, up to 10,000 characters. Write in the language you want to hear; Hebrew works.
2. Generate: The estimated cost in credits shows before you start. Click Generate speech. When Your recording is ready, listen and click Download MP3.
Under the text, Subtle expression in your voice offers buttons that add reading directions where the cursor is (warm, excited, slow, a whisper and more). Add natural delivery with AI suggests a reading of your text; review it and click Apply to script, and you can always restore the original. Advanced voice settings change the variation and how close it stays to your sample. Punctuation makes pauses; phonetic spelling or Hebrew niqqud can fix a word that is read wrong.
Limit | Value
Voices in your library | Up to 5. Delete one to add another.
Sample used to make the voice | 10 seconds to 3 minutes, from a file of up to 50 MB
Text per recording | Up to 10,000 characters
Recordings kept on the page | The latest ones only. Download to keep a copy.
Privacy: Your file is processed for a moment on the platform’s server, and only the part you chose is sent to ElevenLabs to make the voice. Deleting a voice here deletes it at ElevenLabs too. Recordings you downloaded stay with you.
Q: How much does a recording cost?
A: It is paid in credits by the length of the text. The screen shows the estimate before you generate. If a recording fails, the credits held for it are released.
Q: Can I make a voice of someone else?
A: Only with that person’s explicit permission. You confirm it before the voice is made.
Q: It says Voice verification required.
A: ElevenLabs asked for an extra check of that voice. Contact support to finish it; until then the voice cannot be used.
Q: It says voice creation is not active yet.
A: The platform’s voice service is not connected at the moment. Nothing is uploaded until it is.
Q: Can an avatar or an agent speak in my voice?
A: Voice Studio makes recordings you download. Talking video avatars have their own voice setting, see Talking avatars.
Related: Talking video avatars, Video Studio, Track your credits and earn free ones, Make ad images