AI audio

Generate speech with artificial intelligence

The AI Audio block turns text into natural-sounding speech with AI.

How it works

Pick a voice

Choose from available voices or create your own

Provide text

The text you want turned into speech

Generate audio

The generated audio is stored for use in other blocks

Settings

Text to speak

Choose the text to convert to audio. Multiple languages are supported.

Voice selection

Pick the voice you want from the available list.

Text preprocessing tips

Normalize numbers in text

Recommended: before generating audio, run the text through AI to expand digits into words in the target language.

Example prompt:

"Rewrite every digit as words. Do not mark changes unless they were already marked."

Problem

The TTS stack may mis-detect language and read numbers with the wrong pronunciation

Example: "0.1" → "zero point one" instead of the correct wording in your locale

Fix

Preprocessing turns digits into the right spoken form

Example: "0.1" → "zero point one" spelled out consistently for your language

Output

The finished audio is stored and can be used in other workflow blocks.

Benefits

High quality

Natural-sounding speech powered by ElevenLabs

Custom voices

Use your own voices or pick from the library