# AI Text-to-Speech

AI Text-to-Speech (TTS) converts the text you type into lifelike spoken audio. Instead of recording a message or uploading an audio file, you write the greeting you want and the system produces the prompt for you. This page walks through generating greetings with AI TTS in the Cloud Voice App.

## Where you can use it

You can generate greetings with AI TTS for two features:

- **Call forwarding**: a greeting played before a call is routed to its forwarding destination. Set it under **Preferences > Presence > Call Forwarding**.

  ![The call forwarding greeting field with a Text-to-Speech option](/images/pbx/extension-call-forwarding-greeting.png)

- **Voicemail**: the greeting callers hear when they reach your voicemail. Set it under **Preferences > Voicemail > Voicemail Greeting**.

  ![The voicemail greeting field with a Text-to-Speech option](/images/pbx/ext-vm-tts.png)

## Before you begin

- Your Cloud Voice platform must be running version 84.20.0.128 or later. IZT keeps the platform up to date, so this requirement is normally already met.
- Your administrator must enable the **AI Text-to-Speech** feature for your Cloud Voice platform before the option appears.

## Generate a greeting

1. Open the menu for the feature you want a greeting for (call forwarding or voicemail), then click **Text-to-Speech**.
2. In the window that opens, fill in the prompt details.

   ![The prompt settings window with fields for name, language, voice, and text](/images/pbx/ai-greeting-setting1.png)

   - **Name**: a label that helps you recognize the prompt later.
   - **Language**: the language the prompt is spoken in.
   - **Voice**: the voice profile used to read the text.
   - **Text**: the words to convert into audio.

   :::note
   Write the text in the same language you selected under **Language**.
   :::

3. Optional: fine-tune the audio and preview it.

   :::tip
   Click **Listen** and preview the prompt before you save. Editing the text or nudging a slider and regenerating only takes a moment, so it is worth getting the greeting right before callers hear it.
   :::

   a. Drag the **Speed**, **Volume**, and **Pitch** sliders to the values you want.

      :::note
      **Pitch** is only available when **Language** is set to **Arabic**, **English (Wales)**, **Icelandic**, **Romanian**, **Russian**, or **Welsh**.
      :::

      ![The speed, volume, and pitch sliders for tuning the generated audio](/images/pbx/ai-greeting-setting2.png)

   b. Click **Listen** to generate a preview. The button reads **Generating** while the audio is created, then changes to **Play** once it is ready.

      ![The Listen button before the preview is generated](/images/pbx/ai-greeting-listen-1.png)

      ![The button showing Play after the preview finishes generating](/images/pbx/ai-greeting-listen.png)

   c. Click **Play**, then click ![play icon](/images/pbx/play-f.png) in the pop-up to hear the prompt in your browser.

      ![Playing the generated prompt in the browser pop-up](/images/pbx/play-tts-prompt.png)

4. Click **Save**.

## After you save

The generated prompt is stored and ready to apply to the feature you created it for.

You can also review and organize your saved prompts under **Preferences > Voicemail > Voicemail Greeting > Greeting Management**. For details, see [Manage Personal Voicemail Greetings](/pbx/web-client-user-guide/manage-personal-voicemail-greetings/).

![The greeting management list showing saved prompts](/images/pbx/ai-greeting-management.png)