Speak instead of spell — your words come out as clean text. And when you type, a neural autocorrect fixes the spelling for you. No more re-reading every line. 100% on-device, private, and free.
For a lot of people with dyslexia, the hard part of writing isn't having ideas — it's the spelling, word retrieval, and slow, effortful typing that get in the way of getting those ideas down. Dictation flips the process: you say what you mean, and the words appear.
But plain speech-to-text only solves half the problem. If the transcription is rough or the autocorrect "fixes" the wrong words, you're back to editing line by line. The unlock is pairing accurate dictation with an autocorrect that actually understands context.
Tap the mic and talk — your speech becomes clean text, and Pro dictation styles can match the tone you want. Prefer to type? An on-device neural autocorrect fixes spelling and typos in context, with an optional Pro cloud Ghost Editor for deeper cleanup.
Autocorrect and predictions run 100% on-device with CoreML — no keylogging, no analytics, no ad IDs, no sign-in, and an offline dictation mode. The optional Ghost Editor and dictation styles are a clearly-labeled, opt-in cloud tier.
Adjustable keyboard size, an optional number row, and predictions that surface the next word before you reach for it.
Yes — it removes the spelling and word-retrieval bottleneck by letting you speak instead of type. Pairing it with a good autocorrect means the words land correctly without manual cleanup.
One that combines voice typing with a context-aware neural autocorrect, works system-wide, and runs on-device. That's exactly what MyAIDictation was built for.
Yes. Free to download and use; an optional Pro tier adds cloud cleanup, all dictation styles, and themes.
Autocorrect and prediction are on-device. No tracking, no account, with an offline dictation option. (The optional Ghost Editor and dictation styles are opt-in cloud.)