How to Generate Voice

Text to Speech in Studio

1. Open VocalLab Studio and click Text to Speech in the sidebar, under Create
2. Pick a Voice — browse the voice library or use one of your own clones and designs
3. Type or paste your text in the script box
4. Adjust Voice settings if needed: Model, Emotion, Speed and Expressiveness (VocalLab Studio models use Delivery — Stable, Balanced or Creative — instead of Expressiveness)
5. Click Generate Speech (or press Ctrl/⌘ + Enter)
6. Play back the audio, then click Download MP3

Before you generate, the button area shows roughly how many points the script will cost.

Changing the pitch or tone of a voice

VocalLab doesn't have a pitch slider. The generation settings are Speed (how fast the voice talks) and Expressiveness (how dynamic the read is — higher is more animated, not higher-pitched). To get a deeper or higher-sounding voice:

• Pick a different voice that already has the tone you want — browse the voice library.
• Design your own voice by describing the tone, gender, age and accent you want in Voice Design.
• Clone a voice from a sample that has the tone you're after — see How to Clone a Voice.
• On the VocalLab Studio model, use a steering instruction like [in a low tone] — see Voice Steering.

Longer content and multiple voices

For a long script, or one with several characters, use the Audiobook editor (sidebar → Audiobook). You write chapter by chapter, give each line its own voice, generate, and export the finished project. See How to Make an Audiobook. For two-host shows, use Podcast — see How to Make a Podcast.

Tips for best results

• Keep text blocks under 500 characters for fastest generation
• Use punctuation — commas and periods create natural pauses
• For questions, use a ? — the voice will naturally rise in pitch
• For emphasis, write in ALL CAPS (use sparingly)

Supported content

✅ Any language the voice supports
✅ Numbers, dates, currencies (spoken naturally — Text normalization under Advanced settings is on by default)
✅ Abbreviations (spell out for best results: "AI" → "Artificial Intelligence")
✅ Long scripts — text over 2,000 characters is automatically split into chunks and stitched into one seamless audio, up to your plan's per-generation limit (no need to split manually)