Accessibility statement

ReadAnyText exists so that written text can be listened to instead of read. That makes accessibility the product rather than a feature of it, and this page says plainly where the site currently stands.

What is supported

• Keyboard operation. Every control - the editor, the voice picker, the player, the panels - is reachable and operable by keyboard, with visible focus outlines.
• Screen readers. Controls carry labels, and the things that change on their own - render progress, errors, the allowance countdown - announce themselves through live regions rather than changing silently.
• Contrast. Text and interface colours are chosen against the WCAG 2.2 AA contrast ratios, and the one accent colour is validated for text use rather than assumed.
• Reading support. Playback speed from half to double, pauses you place yourself, pronunciation respellings that change only the audio, and word-accurate captions for following along.
• No time limit on reading. Writing, importing and previewing voices need no account; generating audio asks for a quick sign-in, with Google or an email address, that keeps your script on screen. No page expires while you are using it.
• Reduced motion. Animations are decorative and small; none conveys information on its own.

Known limitations

These are real and current. They are listed because a statement with no limitations is one nobody has tested.

• Scanned PDFs do not work. A scan is a picture of text, so there is nothing to extract and nothing to read aloud. Text recognition is not implemented yet.
• Equations, tables and diagrams read poorly. They were written to be looked at, and a voice reading them linearly produces nonsense.
• The studio is dense on small screens. The console packs a lot into a narrow space, and reaching some panels takes more steps on a phone than it should.
• The site has not had a formal third-party audit. The commitments above come from our own testing, which is not the same thing, and we would rather say so than imply otherwise.
• Transcription needs a capable device. The speech model runs on your own machine, which is what keeps recordings private, but it is demanding on older hardware.

Standard we aim at

WCAG 2.2 Level AA. We are not claiming to have been independently certified against it; we are saying it is the bar we work to and the one we will treat a reported failure against as a bug.

Reporting a barrier

If something here is unreachable, unreadable or fights your assistive technology, that is a defect and we want to hear about it. Please include the page, what you were trying to do, and what you are using - browser, screen reader, magnifier, switch device, or anything else. Write to support@readanytext.com, or use the contact form.

We aim to reply within five working days, and to tell you honestly whether a fix is coming and roughly when - including when the answer is that it is not something we can address soon.