Screen reader view: the page, read aloud in order

At least 2.2 billion people¹ live with a vision impairment, and about a billion of those cases could have been prevented or have not yet been treated.

In WebAIM’s 2026 survey² of 1,780 screen reader users, most were blind, but nearly one in five had low vision, and some had no disability at all: they were the testers and developers checking other people’s work. Nine in ten use a screen reader on their phone. People with dyslexia often use text-to-speech in the same way.

Asked whether the web is becoming more accessible, only a third said yes; nearly a quarter said it was getting worse.

On a long page, two in three screen reader users jump between the headings² first, and nearly nine in ten find the levels useful.

Screen readers keep lists of a page’s headings, links, landmarks and form fields, and people move through them with single keys. So a heading that is only bold text, a link that only says “read more”, or a button that is only an icon vanishes from those lists or turns into noise. An image with no description at all is worse than one marked as decoration: some screen readers read out the file name⁵.

The first screen readers for personal computers came from IBM in the 1980s, led by Jim Thatcher. JAWS followed in 1995, Apple built VoiceOver into the Mac in 2005, and in 2006 Michael Curran began NVDA, a free screen reader that is now one of the most widely used.

Every year WebAIM tests the home pages of the top million websites. In 2026, 95.9% failed⁴, with an average of 56 errors each.

Six kinds of problem made up 96% of them: low-contrast text, images with no description, form fields with no label, links with no text, buttons with no name, and pages that do not say what language they are in. All six are easy to fix, and all six are what this tool and the Good Citizens check look for. Pages that used more ARIA, the attributes added to make custom controls accessible, had more errors, not fewer.

This tool reads the page’s structure, so it needs real text, headings and buttons. A Figma prototype is drawn as a picture, with nothing underneath to read, and the tool will say so rather than report nonsense.

It does work on prototypes built as real pages: Axure, UXPin and Justinmind exports, Framer and Webflow previews, Storybook stories, and apps generated by tools like Figma Make or v0. If the content sits in a frame from another site, the tool offers to open the frame on its own.