# Enable AI Text-to-Speech

AI Text-to-Speech (TTS) converts text you type into spoken audio, so you can build voice prompts (greetings, menu announcements, hold messages) without recording anything yourself. It is turned off by default. This page covers the one-time setup that switches the feature on for the whole system. Once it is on, TTS becomes available wherever prompts are used, so you only have to do this once.

## Before you begin

Confirm that your deployment meets both of these conditions:

- **Firmware**: version 84.22.0.138 or newer.
- **Subscription**: an active Enterprise Plan or Ultimate Plan.

:::caution
Both requirements are hard prerequisites. If the firmware is older than 84.22.0.138, or the account is on a plan below Enterprise, the Text-to-Speech toggle will not appear (or will not save). Check and update these first, otherwise the steps below will not work.
:::

## Turn on AI Text-to-Speech

1. Sign in to the PBX (Private Branch Exchange, your phone system) web portal, then open **AI > AI Toolbox**.
2. Enable the **Text-to-Speech** toggle.

   ![The AI Toolbox panel with the Text-to-Speech option switched on](/images/pbx/enable-ai-tts.png)
3. Select **Save**.

:::note
The **AI Toolbox** page (under the **AI** menu) is the master control for all AI features. The **Text-to-Speech** toggle here is a system-wide switch: turning it on makes TTS available to every user and every prompt-based feature. Your change does not take effect until you click **Save**.
:::

AI Text-to-Speech is now active and ready to use.

## Next step

With the feature enabled, you can start creating spoken prompts from text. See [Generate Audio Prompts using AI Text-to-Speech](/pbx/ai-guide/generate-audio-prompts-using-ai-text-to-speech/).