Voice Studio turns text into speech. You type or paste what you want to hear, pick a voice, and get a clip in seconds. It is open to every plan, and it is paid for with credits.
Open Voice Studio from the sidebar. It has these tabs: Create, Voices, Library, Studio Apps and Wallet.
The cost of a clip is shown before you generate it, in characters, credits and dollars. Nothing is charged until you click Generate.
Create
On Create you choose how to start:
- Type it. Type or paste your text. The estimate under the box updates as you write.
- Write it for me. Describe what you want ("a 30-second birthday message for my sister, cheerful") and an AI writes a script for you to edit. Writing is free, up to 50 scripts a day. You pay only for the speech.
Then pick a voice, a Quality and a Speed, and click Generate:
| Quality | What you get |
|---|---|
| Phone | 8 kHz, made for calls |
| Standard | 24 kHz MP3 |
| Studio | 24 kHz WAV |
Every quality costs the same: 1 credit per character, in every language and with every voice. Speed changes how fast the voice talks, never the price.
Not sure what to say? The first time you open Create you will see ideas to start from, such as a birthday message, a bedtime story or a radio ad.
Retakes and edits
- Retake a line to hear it again with a different delivery. Each line gets 2 free retakes within 24 hours.
- Edit script changes only the lines you touch. Only the changed and new lines are made again and charged.
- Making the exact same clip again within 30 days is free: it comes from storage, and the receipt says so.
- A clip that fails is never charged.
The multi-voice format
One script format works everywhere: here, in the Voice API, in Podcast from Anything and in Audio Merge.
[Narrator] It was a dark and stormy night.
[Maya] Did you hear that?
[Leo] (slower) Hear what?
(pause 2s)
[Maya] (whispering) Exactly. Nothing.
[Name]at the start of a line sets the speaker until the next tag. A line with no tag continues the current speaker. You can use up to 20 speakers.- Directions that are performed:
(pause),(pause 2s),(slower)and(faster). - Notes. Any other direction, like
(whispering), is kept for you but not performed yet. - Free: speaker tags, directions and pauses. Charged: only the spoken words, 1 credit per character.
Use Cast voices for me to give every speaker a voice, and change any of them from the Cast panel. If you paste a screenplay, a chat log or lines like Name - text, Voice Studio converts it to this format and shows every line before and after. You can Accept the result, Undo it or Convert again.
Voices
The Voices tab has three lists: Featured, My voices and All voices. Search by name, language or style, filter by language, gender, age or style, and play a sample of any voice. Samples are free.
You can also compare voices: pick up to a few voices and hear your own text in each, side by side. This is charged as speech, per voice.
Make a voice of your own from the same tab:
| Option | What it does | Cost |
|---|---|---|
| Clone a voice | Record or upload 30 to 60 seconds of one speaker. You hear it free before you pay. | 50,000 credits (about $5) |
| Design a voice | Describe a voice in words and choose from 3 options. | The characters the samples speak, about 450 credits, then 50,000 credits to keep one |
| Bank my voice | A guided session for anyone who may lose their voice. See Voice Banking. | 50,000 credits |
A voice is yours to keep, with no monthly fee, and it works in the form builder, in every Studio App and in the API. A new clone includes one free re-record within 7 days. Creating a voice needs a first credit purchase and a verified email, and you confirm that you have the right to use the voice you clone.
You can delete a voice from My voices. Voice Studio shows where it is used (forms, phone lines) and asks which voice they should switch to. Audio you already made stays in your Library, and credits spent on the voice are not refunded.
Library
Everything you make is saved in the Library, free to play and download. Search titles and text, sort by newest, oldest, most credits used, longest or starred, and filter by the app that made a clip (Create, Voice API or a Studio App). Each clip has Star, Rename, Use again and Delete.
Switch between Personal and your Team library. Audio made for a team lands in the team library, so everyone on the team can play and download it. What you make while Voice Studio pays from a team you chose in your Wallet stays in your Personal library.
Downloads
Open Download on a clip to choose a format. Downloads are always free.
- MP3 (standard) and WAV (studio quality).
- WhatsApp voice note. An Ogg file that plays in a chat with a waveform and a play button. On a phone, the share sheet opens; on a computer, download the
.oggand send it from WhatsApp. - Stems. A ZIP with one file per speaker, for scripts with several voices.
Safety
Text is checked before it is spoken. If it looks like sexual content involving minors, a threat, hate speech, a scam, false information about an election or someone pretending to be a real person or organization, it is refused and you are not charged. After several blocked requests Voice Studio pauses for a while. If you think a refusal is a mistake, contact support.
What it costs
Speech is 1 credit per character. The full list, including the Voice API rate, is the rate card. Credits are taken when you click Generate, never at the end, and they come back automatically if we fail to deliver. See Credits.
Related
- Studio Apps: ready-made tools built on your voices
- Voice API and MCP Integration: use Voice Studio from code
- Telephony: put a voice on a phone line
- Text-to-Speech: read your forms aloud