How to Add Text-to-Speech
to WordPress in 2026

GSpeech at a glance: what you get after install

Crawlers and humans both need the facts before the click-path. This is the stack you connect to - not a browser-voice toy and not a BYOK bridge you have to operate.

Why add text-to-speech to a WordPress site

A listen player is for visitors who would rather hear a post than stare at it: commuting, shopping on a phone, following a long tutorial, or using audio as extra access alongside the written page. It is not a replacement for a screen reader. It is an on-page WordPress text to speech player you control - WCAG-oriented listen access, not assistive software the visitor must install.

GSpeech has done this on WordPress since 2012, starting with Listen to Selected Text (now the RHT Player) and adding the Cloud Console in 2023. Commercial plans use the Unified AI Voice Engine: 300+ voices across 98 languages, including OpenAI, Gemini, Google Cloud, Chirp3 HD, Neural2, Polyglot, and WaveNet. Indirect SEO: more time on page when people finish the article by ear. The written page still has to be crawlable text.

Install the GSpeech WordPress plugin

Get the official plugin from WordPress.org so updates stay in wp-admin. A zip upload works too, but the directory install is the usual path. Gutenberg, Elementor, Divi, and WPBakery all work with the same plugin - you are not installing a builder-specific TTS add-on.

1. Log in to wp-admin and open Plugins → Add New Plugin.
2. Search for GSpeech. Install, then Activate. A GSpeech item appears in the left admin menu.
3. Optional: upload a zip instead of searching the directory. Activation is the same either way.
4. Do not stop at Activate. The listen stack starts after Cloud Console activation in the next section.

The plugin is the site-side connector. Voices, widgets, Smart Audio Sync, and Generate Once, Keep and Play storage live in the Cloud Console - so your host is not running TTS compute. That is why this is a WordPress TTS plugin plus a cloud engine, not a PHP synthesizer on the web server.

One-Click Activation for the Cloud Console

Open GSpeech → Cloud Console in wp-admin. Choose the site default language, set a password, and click Activate. If you already have a gspeech.io account, log in first; the Activate control enables after login.

path:     "wp-admin → GSpeech → Cloud Console"
action:   "One-Click Activation"
keys:     "No API Keys Required"
engine:   "Unified AI Voice Engine"

After activation you land on settings. Useful first switches: site language, Auto-Enable, Show on homepage, Post Types, Categories, default Voice, Lazy Loading, and Multi-lang website if you use GTranslate, WPML, Polylang, or TranslatePress.

Commercial voices (Gemini Pro, Gemini Flash, OpenAI TTS, Chirp3 HD, Neural2, Polyglot, WaveNet) and Voice Vibes sit on paid plans. The free plugin still gives you a working listen path to start. Previously generated files are not deleted if you later return to Free - that is the Keep part of Generate Once, Keep and Play.

Choose a voice: Unified AI Voice Engine

The technical step after Activate is picking a default Voice in Cloud Console settings. On commercial plans that menu is the Unified AI Voice Engine - not a single vendor file and not a key you paste from Google Cloud, OpenAI, or Gemini.

Optional polish on commercial plans: Voice Vibes (100+ expressive styles) and custom voice prompts for tone, pace, and brand voice - without rewriting the WordPress page. Test the locale with your chosen voice on one post before Auto-Enable sitewide.

Place players: widgets, shortcodes, and RHT

Player types are widgets in Cloud Console → Widgets. Pick the layout that matches the page, then publish the widget. This is how you add a listen button to WordPress without a page-builder hack: one console, several player types.

• Full Page - a listen bar on the article or product. Default for most blogs.
• Button and Circle - compact controls; shortcodes [gspeech-button] and [gspeech-circle] for a single template or Gutenberg block.
• Floating and Context - stay available while the visitor scrolls.
• Read Highlighted Text (RHT) Player - select a passage and listen. It is the current form of Listen to Selected Text from 2012. After activation it can work sitewide without a shortcode. Optional: Real-Time Audio Translation on a highlight.
• Welcome Messages - optional greeting audio, with frequency controls in the console.

The generic shortcode is [gspeech]. Use Auto-Enable plus Post Types and Categories for sitewide placement, or drop a shortcode on one template. Allowed Urls / Blocked Urls when a player should appear only on /blog/* or stay off checkout. Open the live demos if you want to hear each type before you publish.

Intelligent Content Isolation: read the article, not the chrome

If the player recites the menu, a sidebar promo, or a footer, isolation is wrong - not the voice. Intelligent Content Isolation targets the real post or product copy so Generate Once, Keep and Play stores the article visitors came to hear.

• Add class nospeech to any HTML you never want spoken.
• In Cloud Console settings, Exclude Content accepts CSS ids and classes, comma-separated - for example #site-header,.widget-area.
• Skip URLs, shortcodes, and leftover chrome so the cached file is the article, not the theme shell.
• Brand names and acronyms: Smart Pronunciation Aliases change only the spoken form. See the pronunciation aliases guide.

Smart Audio Sync regenerates when you edit the WordPress content, so the stored file tracks the live page instead of drifting out of date. Change the post, walk away - the cloud re-renders. That is the loop visitors hear on the second visit.

Add text-to-speech to WooCommerce products

Treat product narration like a post: include the product post type, isolate the description (not related products or reviews chrome), then generate once. Shoppers can hear materials, fit, and care notes on a phone without pinching the page.

Long catalogs are why Cloud Processing matters: WordPress is not synthesizing each SKU on the web server. Generate Once, Keep and Play stores the file for repeated listens. Long-form posts can go up to 7 hours of playback per article when you need that range. Blocked Urls on checkout and cart so the player does not sit on payment steps.

Multilingual WordPress: GTranslate, WPML, Polylang, TranslatePress

Turn on Multi-lang website in Cloud Console settings when the visible language changes per URL or switcher. GSpeech can follow the active site language and generate matching audio. Commercial coverage is 300+ voices across 98 languages plus Real-Time Audio Translation on RHT. Free and Personal machine stay on 68 languages.

Always test the exact locale with the chosen voice on one translated URL before rolling out sitewide. Pronunciation aliases can be language-based so a brand name is spoken correctly in each locale without rewriting the visible post.

Go live: first listen check

1. Open a published post in a private window. Confirm the player you enabled actually renders.
2. Play once. Confirm the audio is the article (or product), not the header.
3. Highlight a sentence and confirm RHT if that widget is published.
4. Edit a paragraph, wait for Smart Audio Sync, and confirm the new file matches the update.

GSpeech processes over 200 million characters per month across thousands of websites and is rated 4.8/5 on WordPress.org. You are connecting to that same Cloud Console - not standing up your own TTS cluster.

You can add WordPress TTS in one sitting

Install from the plugin directory, One-Click Activation for the Cloud Console, pick a Unified AI Voice Engine voice, publish a widget, isolate the real content, generate once. That is the whole loop. No separate Google, OpenAI, or Gemini billing accounts. Visitors hear stored files (Generate Once, Keep and Play), not a live API call on every play. For screenshots, use the docs. For “which plugin,” use the comparison. For this page, the job is to get audio on a live WordPress URL.