Introducing Text-to-Speech Voice AI

We’re thrilled to announce that Text-to-Speech Voice AI is now available on Appaca Studio! You can use models from OpenAI and ElevenLabs to power your app’s Voice AI.

Select model and voice

Both the OpenAI TTS model and the ElevenLabs model offer amazing voices. You can choose from the available voices to power your Voice AI model.

Text-to-Speech Action

You can use your voice model in an Action by selecting the Text-to-Speech action in the action browser. Simply provide the text you want the AI to speak. This is especially useful when combined with a text model—for example, the text model can generate a story, and the voice model will read it out loud.

New audio state variable

We've also shipped a new state variable to store the audio-type values. The text-to-speech AI action will respond in audio-type value. This value can be stored in this new audio state variable.

Audio component

We've also added a new Audio component. With the audio component, you can play audio by simply passing the audio-type value from your state variable to the audio source.

Play audio with Interaction

If you want to play the audio immediate right after the action is triggered for text-to-speech, you can use the new interaction called "Play audio". You can pass the audio-type value from the state variable to the Audio field. It will automatically play the audio as soon as the audio-type state variable is available.