Skip to main content
Innovations

aXread

Active

Read Any Website Your Way

AI-powered reading overlay that transforms complex web content into easy-read format. Real-time simplification, visual aids, and text-to-speech with word-by-word highlighting.

The insight: What if we could give people with cognitive disabilities the same independence everyone else takes for granted when reading websites?

Opens the aXread panel on this page

The Problem

My son has complex disabilities including intellectual disability. Every website is a barrier.

What Exists

  • xScreen readers announce text literally—doesn't help if the text is too complex
  • xBrowser read-aloud uses original content—bureaucratic language stays bureaucratic
  • xEasy-read documents require manual creation—most sites don't have them
  • xRelies on someone else to read and explain—loss of independence

What aXread Does

  • +AI simplifies content before reading—matches your comprehension level
  • +Works on any website—no special version needed
  • +Visual aids (emojis) help track position—hidden from TTS
  • +Independence—access information without waiting for help

Before & After

Real NDIS text transformed to easy-read format.

Original

"The National Disability Insurance Scheme (NDIS) provides funding for individualised supports and services other than day-to-day living costs to eligible people with permanent and significant disability."

aXread Version

The NDIS gives you money for help.

It is for people with disabilities.

You choose what help you get!

Key Features

AI Text Simplification

Gemini Flash transforms complex content into plain language. Configurable reading levels from age 5 to 12.

Visual Emoji Anchors

Emojis added as visual cues to break up text. Hidden from screen readers with aria-hidden so they're seen but not announced.

Synchronised TTS

Text-to-speech with word-by-word highlighting that tracks exactly where it's reading. Auto-scrolls to keep pace.

Page Highlighting

Original content blocks highlighted on the page as they're read. Maintains orientation in complex documents.

Tone Selection

Choose how content sounds: friendly, neutral, fun, serious, or sarcastic. Personalised reading experience.

Multi-Provider AI

Supports OpenAI, Gemini, VLLM, or custom endpoints. Self-host with your own models if needed.

Architecture

Built as a modular React component system. Drop it into any Next.js application or run standalone.

Content Extractor: Parses DOM to identify readable content blocks (headings, paragraphs, lists)

AI Transformer: Sends content to LLM with reading level and tone prompts

Speech Highlight Hook: Web Speech API with onboundary events for word tracking

Overlay Panel: React component with settings, navigation, and TTS controls

LLM Config: Environment-based provider selection with cost tracking

Technical Specs

Default AI
Gemini 2.5 Flash Lite
Cost Per Transform
~$0.00008
Response Time
<1 second
Framework
React / Next.js
Speech API
Web Speech Synthesis
Highlighting
Word-index tracking

Why This Matters

Independence isn't just about physical access. It's about information access. Every time someone with cognitive disabilities needs to wait for a support worker to explain a website, that's a loss of autonomy.

aXread isn't just a tool—it's about dignity. About being able to research your own healthcare options. About understanding your own NDIS plan. About not being dependent on someone else's availability to access information.

Neurodivergent-led innovation means building from lived experience, not assumptions about what people need.

Want aXread For Your Platform?

aXread can be integrated into websites, applications, or service portals. Make your content accessible to everyone.