A personal voice from eight seconds of audio
Dawn includes 970 natural, neural voices across ages and accents. It can also build a new voice from about eight seconds of recorded speech, because that's what the engine was trained on. The voice is created, measured against the recording, and tuned on the device in about ten seconds.
That can be the user's own voice, preserved while speech is still available, or a family member's. The recording and the voice stay in the app. Dawn has no server and no account system, so there is nowhere else for them to go.
The Voice Studio holds all 970 voices. Shape any of them by pitch and age, blend two into something new, or hand Dawn a recording and let it surface the voices nearest to it. One take is enough, even a shaky or noisy one, which matters when a steady voice is exactly what someone no longer has.
Message banking is also built in: record whole phrases and play them back exactly as they were said.
Eight seconds in → their voice out. All on the device.
No lag between the tap and the voice
Dawn synthesizes the message in the background while it's being composed, and keeps the words on the board pre-rendered for the active voice. Tapping Speak plays immediately.
Words are also checked before they're cached: the engine verifies its own pronunciation against the written word and re-renders until it matches, so "doll" never comes out "dolls".
Write it, hear it
Dawn has a built-in whiteboard for the moments when drawing is faster than finding the word. When what's on the board is writing, one tap reads the handwriting, shows what it read, and speaks it in the selected voice.
It's useful in both directions. A communication partner can write two choices on the screen and have Dawn ask the question out loud. Handwriting recognition runs on the device, like everything else in the app.
Fix a tense with one tap
Tap a word and its whole family appears next to the board: am, are, is, was, were, the contractions, and related verbs. Changing "is" to "was" is one touch, with no digging through folders.
The board itself never rearranges. Every word keeps the same position on every grid size, on iPad and on iPhone, so practiced motor patterns keep working.
Everything you'd expect from a full AAC app
3,400+ symbols
The complete Mulberry symbol set, plus personal photos as symbols. A guided picker and search that answers "where is that word?"
Switch access & scanning
Row–column scanning with external switch support, recoverable from every state, verified by automated lockout testing.
Full customization
Edit any button's word, symbol, and color. Add words and folders. A typed-math caregiver gate keeps editing away from busy hands.
Typing with prediction
A large, always-visible keyboard with word prediction and caret controls, for spellers, with the same instant voice.
Three grid sizes
4×6, 6×8, and 8×10 layouts where the smaller grids embed exactly into the larger, so a growing vocabulary never relearns motor plans.
Open formats & backups
Import Open Board Format boards from other systems. One-tap full backups that restore everything, kept in your own files.
Touch accommodations
Hold-to-activate timing, VoiceOver support, and Dynamic Type. Access needs differ, and defaults never assume.
iPhone and iPad
One purchase, both devices, the same board layout. The backup device in a parent's pocket can be the primary in a pinch.
Whiteboard
Six pens and an eraser for the conversations that need a drawing, and it can speak what's written on it.
How Dawn compares
Against what leading symbol-based AAC apps typically offer.
| Dawn | Typical premium AAC | |
|---|---|---|
| Only on Dawn | ||
| Personal voice from an eight-second sample, created on the device | Yes | Not offered, or cloud-based |
| A Voice Studio: 970 voices to shape, blend, or match to a recording | Yes | A short fixed list |
| Modern neural voices with no internet, ever | Yes | Older voice tech, or online features |
| Handwriting on the whiteboard, spoken aloud | Yes | Not offered |
| No account, no sign-in, nothing collected | Yes | Varies |
| The fundamentals, in both | ||
| Symbol boards with full customization | Yes | Yes |
| Switch scanning and alternative access | Yes | Yes |
| Typing view with prediction | Yes | Yes |
| Grammar support | Yes | Yes |
| Message banking | Yes | Often |
Common questions
Is the voice recording safe?
The recording is processed into a voice on the device and stays in the app's storage on that device. Dawn has no server, no account system, and no analytics. Deleting the voice deletes it.
Does it need the internet?
No. Not for setup, not for speech, not for the personal voice. Everything ships in the app. Airplane mode changes nothing.
Does the whiteboard need an Apple Pencil?
No. Finger drawing is first-class everywhere in Dawn, since most AAC users don't have a Pencil, so nothing assumes one.
What devices and languages?
iPhone and iPad on iOS 17 or later, one purchase for both. English at launch; the symbol set and speech engine are built with more languages in mind.
Who is Dawn for?
Nonspeaking and minimally speaking people of any age: autistic children and adults, people living with ALS, aphasia after stroke, apraxia, Parkinson's, or cerebral palsy, along with their families, SLPs, and care teams. Dawn covers the range from first symbols through full sentences and typing.