accessibility reader
Make any content readable aloud with natural voices your users trust.
- Sub-300ms first-byte streaming
- 1,500+ natural voices, 30+ languages
- Speech marks for word-highlight sync
- From $6 per 1M characters
definition
An AI accessibility reader speaks written content aloud: articles, documents, app screens, and web pages. Built on text-to-speech, it gives readers with dyslexia, low vision, or a preference for audio a natural voice and a follow-along experience, rather than a robotic screen reader.
natural-voices
The voice is the difference between a feature people use and one they turn off. With 1,500+ natural voices across 30+ languages, users pick a voice they trust and can listen to for a long session without fatigue.
word-highlighting
Speech marks return word-level timing, so the interface highlights each word as it is spoken. That follow-along highlight helps readers track the text and is a core accessibility affordance for dyslexia and low vision.
fast-start
Streaming returns the first audio in under 300ms, so playback begins almost immediately on press. A reader is not left waiting for a full render before hearing anything.
pricing-note
Per character, from $6 per 1M on Scale. See the pricing page.
Frequently asked questions
Can I add a read-aloud feature with text-to-speech?
Can I highlight words as they are read?
How quickly does playback start?
Start building
Make any content readable aloud with natural voices your users trust.