Hello Paul,
Thank you for your reply and for taking the time to consider my suggestion.
However, I think there may be a small misunderstanding about the main purpose of this feature.
AI Voice on Hover is not intended to be an accessibility tool or an alternative to existing screen readers.
It is primarily about creating a completely new interactive browsing experience for all website visitors, by adding a natural voice dimension directly to website content.
How would it actually work?
Imagine a webpage containing several paragraphs.
- The visitor hovers over paragraph 2: a natural AI voice reads only that paragraph aloud.
- The visitor moves the mouse to paragraph 5: the previous audio stops, and paragraph 5 is read aloud instead.
- Visitors can freely explore a webpage and listen only to the specific content they are interested in.
There is no need to find a Play button, open an MP3 player, or listen to an entire article.
The text itself becomes an interactive, voice-enabled element.
Real-world applications, especially for language schools
This concept could be particularly valuable for tourism, cultural and educational websites, product presentations, immersive experiences, and especially for language schools like ours.
As a language school, we could allow students to hover over a word, sentence, or paragraph and instantly hear its natural pronunciation in the language they are learning.
This would help students directly connect written text with spoken language, improve their listening comprehension, and make learning more interactive, without constantly clicking audio playback buttons.
This is exactly the kind of learning experience I would love to create with SP Page Builder Pro.
Regarding TTS API costs
I completely understand your concern about the cost of AI Text-to-Speech services.
However, I am definitely not suggesting generating audio every time someone hovers over a paragraph!
The idea is to generate each audio file only once, when the content is created or updated, and then cache it on the Joomla server.
Visitors would simply play these existing audio files, without triggering any additional TTS API requests.
If a paragraph is modified, only its corresponding audio file would need to be regenerated.
ElevenLabs could be the preferred provider for its exceptional voice quality and realism, while Voicemaker.in could offer an attractive alternative for voice variety and cost efficiency.
Support for custom MP3 files
I would also suggest allowing users to associate an existing MP3 file directly with a specific paragraph.
This would make it possible to use human-recorded voices, audio generated by other services, or professional narrations without requiring a TTS API.
The interaction would remain exactly the same: the audio would play when the visitor hovers over the corresponding paragraph.
Integration into SP Page Builder Pro
Ideally, the feature could be configured at three levels:
- Global Settings: Choose the TTS provider, voice, language, and default settings.
- Section Level: Enable voice reading for all supported paragraphs within a section.
- Text Block Level: Optionally customize the voice, assign a custom MP3, or disable voice playback for specific elements.
To comply with browser autoplay restrictions, visitors could first activate an optional "Voice Browsing Mode" before using the hover-to-speak interaction.
The real purpose of my suggestion
My intention is not to replace existing accessibility technologies.
I am proposing a new way of interacting with websites: a website where text naturally comes to life and speaks when visitors move their mouse over it.
For us, as a language school, this would be particularly valuable from an educational perspective, but I believe the potential goes far beyond language learning.
I hope this clarification better explains the concept and how it differs from a traditional MP3 player or an accessibility tool.
Thank you again for your attention and for sharing this idea with your development team!