AI Text to Speech

How can we help?

Search the documentation or ask the AI agent anything about the plugin.

Generate audio from any text on the Generator page

Updated September 15, 2026

The Generator page turns any text you paste in into an MP3 file you can play, copy the link to, or download. It’s at AI Text-to-Speech › AI Text-to-Speech (labelled Generator in the plugin header bar) and is useful for podcast intros, video voiceovers, announcements, or testing voices before you generate audio for real posts.

Only administrators (users with the manage_options capability) can open this page. The controls are greyed out until an OpenAI API key is saved in the settings.

Create an audio file

  1. Type or paste your text into Enter your text here. The character count and estimated cost update as you type.
  2. Pick a voice under Select a voice.
  3. Pick a style under Select a voice style. This only appears when the voice model is GPT-4o mini TTS.
  4. PRO users can add Custom Voice Style Instructions, such as an accent, pace or tone.
  5. Click Generate TTS Audio and wait for “Generating audio, please wait…” to finish.

The result shows the text you converted, an audio player, the file URL with a Copy button, a Download Audio File button and Generate New Audio to start again. Each generation sounds a little different, so it’s worth trying again if you don’t like a take.

Where the files are kept

Files are named admin- followed by a random string, for example admin-8poJsikOQk.mp3, and saved to your chosen storage location. They’re listed on AI Text-to-Speech › TTS Files as “Admin Generated”.

The plugin keeps a record of the 50 most recent Generator files. When you create the 51st, the oldest entry is removed and, if it was stored locally, its file is deleted from the server. Download anything you want to keep permanently.

What gets applied to Generator text

  • The same clean-up as post content: HTML tags, image tags, URL prefixes and quote marks are removed, and currency amounts are written out as words.
  • With PRO: global Custom Text Before Content and Custom Text After Content, custom pronunciations, and the [voice], [tts say] and [tts ignore] shortcodes.
  • The post title, author and date options don’t apply, because there’s no post.

The Generator page doesn’t attach audio to a post. To add a player to a post, use the AI TTS box in the post editor.

Related

Was this article helpful?