Skip to main content
Voice Studio is where you work with voices. Write a line, choose who says it, and hear it — then save the voices you like with your own settings, or clone one of your own from a short recording and use it across your Pearls. Open it from Voice Studio in the left sidebar. The page, titled Text to Speech, has three parts: the draft on the left, where you write; the settings on the right, where you choose the voice and adjust it; and the generations below, where your takes are listed.
The Text to Speech page, with the draft, the voice settings and the generations
The Text to Speech page, with the draft, the voice settings and the generations
A voice is a speaker in a language, so choosing a voice already chooses the language it speaks. That’s why there is no separate language selector on this screen.
Voice Studio requires the TTS permission (Access Voice Studio - clone voices and generate speech). See Users & Workspaces to manage permissions.

Generating speech

1

Choose a voice

At the top of the settings, click the current voice to open the voice library and pick another one. See The voice library. Click the flag of the voice to hear a sample of it. Voice Studio remembers the last voice you used.
2

Write what the voice should say

Type into the draft. A single take can be up to 4,000 characters. Past that limit, the page tells you by how much (Over limit by … characters) and Generate speech stays disabled until you trim it.On an empty draft, Start with one of these offers a few ready-made lines: click one to use it. Once the draft has text, the Example line button (the shuffle icon) fills it with a line written to show the selected voice off: hesitations, a breath, a number said out loud. Press it again for a different one.Type @ anywhere in your text to insert a speech marker — a short tag the model reads as an instruction rather than as words. See Speech markers below.
Clean sentences sound the same in every voice and tell you very little about the one you picked. The example lines are written to bring out what makes a voice different.
3

Generate and listen

Click Generate speech. Playback starts as soon as the first audio is ready — you don’t wait for the whole take to finish generating, which matters for longer texts.Each take is added to the generations under the draft, newest first. For each one you see the text, the voice, the time and the duration. You can:
  • Play and pause it with the flag at the start of the row, and click along the bottom of the row to jump to a moment.
  • Download the audio.
  • Use this text to put its text back in the draft.
Voice Studio keeps your last 20 takes.

Adjusting a voice

The settings on the right apply to the voice you chose:
The voice settings with the Speed menu open
The voice settings with the Speed menu open
A setting you change shapes the takes you generate right away. To keep it, save the voice:
  • For a voice of the library, click Save as my voice, give it a Voice name, and click Save. The voice is added to My voices with your settings.
  • For a voice of My voices, click Save changes to update it, or Save as new to keep it as another voice of your list, with its own name.
Save as my voice, with the name to give the voice
Save as my voice, with the name to give the voice
Your agents use the voices of My voices with the speed saved on them. See Using your voices in a Pearl.

Speech markers

Markers are tags the speech model reads as instructions, not as words. Type @ in your draft to open the marker menu at your cursor, then pick one.
The marker menu open at the cursor, offering Laughs and Breath
The marker menu open at the cursor, offering Laughs and Breath
Only the markers listed above are recognised. Any other bracketed tag you type — including one that looks like it should work — is read out loud, brackets and all. For a beat in the middle of a sentence, use punctuation, or a [breath] mid-clause.

The voice library

Click the current voice at the top of the settings to open the voice library. It has two tabs:

Explore

Every voice available to you, for choosing: the NLPearl voices, and the voices other accounts have published.

My voices

Your account’s own list, for managing: the voices you saved, added or cloned. Your Pearls can be set to these voices.
The voice library on the Explore tab, with search, filters, and a voice card hovered
The voice library on the Explore tab, with search, filters, and a voice card hovered
A voice reaches My voices in three ways: you clone it, you add it from Explore, or you save it from the settings with Save as my voice.

Searching and filtering

Search by name, accent or tag, and narrow the list with the filter chips:
  • Any language
  • Any accent — appears once you’ve picked a language, and only for languages that have accent variants.
  • Any gender
  • Your clones — on My voices, shows only the voices you cloned.

Reading a card

Each card shows the language flag, the voice’s name, and its accent and tags. Some cards also carry a label: Hover a card to Use it (or see that it is In use), to Add it to My voices from Explore, or to open its ⋯ menu. You can also click the card itself to use the voice.

Managing a voice

The ⋯ menu of a card offers actions that depend on the tab and on whose voice it is:
The menu of one of your clones on the My voices tab
The menu of one of your clones on the My voices tab
Remove from my voices and Delete voice both take the voice away from any agent already using it: those agents lose their voice. Delete voice cannot be undone — the voice and its recording are gone for good.

Cloning a voice

You can build a voice from a short recording of someone speaking, then use it like any other voice on the platform. Click Clone a voice in the voice library. Cloning takes two steps: Recording, then Details.
1

Provide one recording

Drop an audio file on the window, click Browse files, or click Record audio to record straight from your microphone. While you record, Need something to read? shows a script to read aloud, in the language of your choice; click Done to stop.
The Recording step of Clone a voice, asking for one recording
The Recording step of Clone a voice, asking for one recording
A voice is built from a single recording. Dropping a second file replaces the first rather than adding to it.
For the best result, use clear, continuous speech with no background noise — one person talking, no music, no overlapping voices.
2

Describe the speaker and confirm your permission

Still on the Recording step, describe the speaker:Then tick the box confirming that the voice in the recording is your own, or that you hold documented written permission from the person whose voice it is, and that you authorize NLPearl to create a synthetic replica of that voice.
The Recording step with a recording, the language, accent and gender fields, and the permission checkbox
The Recording step with a recording, the language, accent and gender fields, and the permission checkbox
Cloning someone’s voice without their explicit, documented permission is not permitted. The confirmation is a legal attestation, not a formality.
Click Next. Your recording is uploaded and the voice is built — this takes a few seconds.
3

Hear the sample and name the voice

On the Details step, the Sample plays a line in the cloned voice, so you know what you’re keeping. This is also the clip the library will play for this voice. If it doesn’t sound right, go Back to change the recording.Give the voice a Name (2 to 25 characters), and optionally up to 3 Tags to make it easier to find — for example warm, support, news.
The Details step, with the sample player, the name and the tags
The Details step, with the sample player, the name and the tags
If your recording contained little actual speech, a warning next to the sample says so: the clone still works, but a longer take makes a closer match.
4

Save

Click Save to my voices. The voice is added to My voices and becomes the selected voice, ready to generate speech.
If a recording can’t be read, try another file or record it again. After many attempts in a short time, cloning pauses for a little while.

Publishing a voice

Publishing puts one of your cloned voices into the shared library, where every account on the platform can use it.
Publishing cannot be undone. Once a voice is published, it stays in the shared library and you cannot withdraw or delete it. Publish only voices you are certain you want to share, and only if the permission you hold for that voice covers it.
Publishing is offered from the voice’s ⋯ menu on the My voices tab, and only while the voice is still private. A confirmation asks you to approve it. Once published, the option is gone and the card reads Published in Explore.

Using your voices in a Pearl

An agent’s Voice dropdown lists the platform’s voices and the voices in your My voices list.
1

Add the voice to your list

Clone it, add it from Explore, or save it from the settings with Save as my voice.
2

Open your Pearl's settings

Go to Pearl Settings → Agent Names, Languages & Voices.
3

Pick it for an agent

Open the agent’s Voice dropdown and select it. For the languages Voice Studio supports, the dropdown also offers Open Voice Studio at the bottom of the list, which opens the studio in a new tab so you can clone or audition a voice without losing your place in the form.
An agent's Voice list in Pearl Settings, with Open Voice Studio at the bottom
An agent's Voice list in Pearl Settings, with Open Voice Studio at the bottom
The agent uses the speed saved on that voice in My voices. Change it from the settings in Voice Studio, then click Save changes.
See Voices and Languages for how agents, languages and voices work together in a Pearl.

Supported languages

Voice Studio generates speech in the following languages:
This is the list of languages the speech model currently serves, which is narrower than the full set of languages a Pearl can speak. See Voices and Languages for everything the platform supports.
Your My voices list is empty until you save, add, or clone a voice.

Voices and Languages

Configure agents, languages and voices for a Pearl.

Pearl Settings

Everything you can configure on a Pearl, including its agents.