Guide›Media Studio
Media StudioBeginner8 min

Voice & speech

Turn text into natural speech inside Myndlab Studio — pick a voice or clone your own, generate across languages, and export the audio for use anywhere.

Text to speech

The voice module in Myndlab Studio converts written text into natural-sounding spoken audio. Paste in text — a single line or a full script — pick a voice, and the Studio generates an audio file that lands in your asset library.

Generation runs on a fal.ai backend, so results are high quality and fast. This is well suited to narration, video voiceovers, IVR prompts, accessibility guides, and audio prototypes.

Choosing a voice

Before generating, choose a voice from the built-in voice library. Each voice has its own character — tone, age, warmth, and pacing. Preview a sample of each voice before committing to it.

1
Browse the voice library
Open the voice module in the Studio and browse the available voices. Filter by gender, style, or language to narrow the options.
2
Preview a sample
Play the audio sample for each candidate to hear its tone and pacing on a standard line before applying it to your own text.
3
Adjust pace and tone
Where available, fine-tune delivery speed and expressiveness to match the mood you want — calm and measured, or lively and energetic.
💡
Tip.For projects with multiple audio clips, stick to one voice across all of them to keep your brand's audio identity consistent.

Cloning a voice

If you want a custom voice — your own, or a spokesperson for your brand — you can clone a voice from a short audio sample. Upload a clean recording, and the Studio creates a reusable voice you can call on in any later generation.

1
Upload a clean sample
Provide a clear recording of the target voice with minimal background noise, music, or echo. The cleaner the sample, the better the clone.
2
Name and save the voice
Give the cloned voice a clear name so you can find it later. It is then added to your voice library alongside the built-in voices.
3
Test it on a trial line
Generate a short line with the cloned voice to confirm it captures the tone and character before you produce long scripts.
⚠
Watch out.Only clone voices you have the right to use. Cloning someone's voice without their permission may violate the law and our terms of service.

Languages & accents

The voice module supports speech generation in multiple languages. Write your text in the target language, pick a voice that handles that language, and the speech is rendered with natural pronunciation and an appropriate accent.

Some voices handle several languages, while others specialize in one. Look for the language indicator on each voice card — such as EN, AR, or ES — to see what it supports.

ℹ
Note.For bilingual scripts, generate each language segment separately with a voice suited to that language, then stitch them together afterward for the cleanest result.

Generating audio

Once you have your text, voice, and language set, run the generation. A new job appears, and when it completes the audio clip lands in your asset library where you can play it, regenerate it, or export it.

text
Voice: "Brand Narrator" (cloned)
Language: English (US)
Script:
  "Welcome to Atlas. Let's get your first project
   set up in under two minutes."

If a clip doesn't sound right, tweak the text — add punctuation to introduce pauses, or split long sentences — and regenerate. Small text changes often improve pacing and clarity noticeably.

Exporting

From the asset library, export any audio clip as a standard file format for use in your video editor, a presentation, or an app. You can also push the audio directly into a Web App or Design project without leaving Myndlab.

💡
Tip.To learn how to push generated audio straight into a live app, see the “Use Studio assets in a project” guide.