Speech

Speech

The Speech tile is designed to convert text into audio using AI voices. You can select the style, voice, speaker roles, and model to get the delivery you want — from a business presentation to a podcast or dialogue.

What does it do?

Turns text into speech. Can voice a specific text or generate speech on a given topic.

When to use?

When you need to quickly get an audio file for a scenario, line, or topic — for example, for a presentation, voice assistant, podcast, or UX prototype.

Main tile elements

Text for speech synthesis

Enter the phrase or topic. You can specify whether the text is generated automatically or provided exactly as written.

Speech style

Choose tone and delivery: neutral, podcast, conversation, presentation, etc. You can also specify the style directly in the text.

Speakers

Configure voices: name, role. Multi-voice synthesis is supported (for example, dialogues).

Model selection

Choose the language model that generates or edits the text for speech. More: Choosing a language model.

“Apply” button

Starts the generation and speech synthesis process.

How to use

1. Create a Speech tile.
2. Enter the text or topic for speech.
3. Set style, speakers, and roles.
4. Set a language model if needed.
5. Click Apply.

Recommendations

• Specify the style for more lively delivery.
• Name and role of speakers are especially important for dialogues.
• For a topic-based dialogue, use scenario-based generation mode.