Skip to main content
The TTS Playground makes it easy to try out Inworld’s TTS capabilities through an interactive playground. It can be used to find the perfect voice for your project, test different text inputs, adjust voice settings, and experiment with audio markup tags. TTS Playground

​
Get Started

1

Go to Inworld Portal

In Portal, select TTS Playground from the left-hand side panel.
2

Enter text

Enter the text you want to convert to speech. If you need some ideas, you can select one of the suggestion chips at the bottom of the screen. Note that character limits in TTS playground depend on your plan, see the Pricing page for details. You can also use our API to generate longer content, see Long Text Input.
3

Select a voice

On the right-hand side, click on the voice dropdown to browse available voices. You can filter by language or search by name, and click the play button next to each voice to hear how the voice sounds. Select a voice.
4

Generate speech

Click the “Generate” button on the bottom right. The audio will automatically start playing once it’s been generated. You can also download the clip to save it.

​
Advanced Features

For greater control over the generated audio, you can try out the following:
  1. Try a different model - Select a different model from the right-hand side panel to see how it compares. See Models for more information about each model.
  2. Adjust configurations - Use the sliders on the right-hand side panel to adjust Temperature and Talking Speed. See here for more information.
  3. Add pauses - Try adding SSML break tags like <break time="1s" /> to insert pauses in your speech. See Pause Controls for more information.

​
Create a Voice

In the TTS Playground, click + Create a Voice to design a new voice from a text description or clone one from audio:
  • Voice Design — Describe the voice you want in text (age, accent, tone, etc.) and get AI-generated voice candidates
  • Voice Cloning — Clone a voice from as little as 3 seconds of audio (up to 15 seconds — longer samples improve similarity)

​
Multi-Voice Generation

Create a multi-speaker composition — a dialogue, scene, or podcast — with a different voice per line, generated as a single stitched track.
1

Add a speaker

Click Add Speaker to add a line. Each line has its own voice, model, and settings, plus inline steering tags like [laugh].
2

Generate all

Enter the text for each speaker and click Generate All to synthesize and stitch every line into one track. Regenerate any line individually to refine it — the combined audio updates automatically.
3

Refine and export

Use the arrows on a line to compare or revert to an earlier take — restoring that line’s text and settings — then play back, download, or share the result.

​
Audio Generation History

Open the History panel using the button next to Generate at the bottom of the playground. Every clip you generate is automatically saved there, capturing the voice, model, and text used. From the History panel you can:
  • Replay any past generation instantly — cached audio plays without re-running the model
  • Reload settings into the playground — send a past entry back to populate the text, voice, model, and all generation parameters in one click
  • Search and filter by text content, voice, model, or date
  • Bulk manage entries — enter Manage mode to multi-select for download or deletion
  • Download individual takes as MP3 with one click
History is stored locally per device and per workspace — it isn’t synced across browsers or shared between teammates. Clearing browser storage will reset it.

​
Next Steps

Ready for more? Whether you’re looking to clone a voice, design one from text, or start building with our API, we’ve got you covered.

Voice Design

Create a voice from a text description—no audio needed.

Voice Cloning

Create a personalized voice clone with as little as 3 seconds of audio.

Best Practices

Learn tips and tricks for synthesizing high-quality speech.

Quickstart

Learn how to make your first API call in minutes.