We aim for WCAG 2.2 AA. Speaking practice should be usable with a keyboard and a screen reader, not just a mouse.
What we do
- Semantic HTML, labelled controls, and a visible focus ring on every interactive element.
- A “Skip to content” link, and landmark regions for navigation and main content.
- Live regions announce the transcript as you speak, and status changes like errors and timer completion.
- Colour is never the only signal — scores and states always have text too.
- Respect for
prefers-reduced-motion, plus a manual “Reduce motion” toggle in Settings. - Large touch targets and a microphone button reachable one-handed on small screens.
- Light and dark themes, both meeting contrast requirements.
- A full typing path for every appropriate exercise, so speech is never the only way in.
Known limitations
- Browser speech recognition quality varies and is outside our control. The typing path is the reliable alternative.
- Reference audio depends on your browser's speech synthesis voices.
Feedback
If something is hard to use with assistive technology, please tell us via the Contact page and we'll prioritise it.