Undo, copy, paste, select —without your hands leaving the keyboard.
AIspeaky ships a set of low-latency, deterministic commands you can trigger with your voice. Text-rewrite commands are expanding — they're shown separately so nothing looks more ready than it is.
AIspeaky is currently in beta and free to use within beta limits. We do not read or store the content of what you dictate.
Two groups — simple and expanding.
Everything is labelled with what it does and how it's triggered, so you don't guess.
- 1
Simple commands
Deterministic actions like undo, copy, paste, select. Low latency and stable in the beta.
- 2
Text-rewrite commands
"Make this shorter", "more formal", "summarize the selection". These call managed models and are marked as expanding.
- 3
Bring-your-own shortcut
Any command has a rebindable trigger. Push-to-talk plus a keyword — no waking words listening in the background.
Simple vs. expanding — never mixed in one grid.
You can see, before you speak, what a command will do and how safe it is.
The exact command families in the beta.
We list what's available and leave the rest visibly out of the promise.
- Undo, redo, copy, cut, paste
- Select word, sentence, line, paragraph, all
- Move caret / extend selection by word or sentence
- Rebindable push-to-talk plus keyword trigger
- General "do anything" agent behaviour
- Application-specific commands beyond deterministic keyboard actions
- Always-on wake word (we don't do background listening)
- Guaranteed rewrite quality across every prompt
Answered plainly.
Is AIspeaky always listening?
No. All commands run inside a push-to-talk session. Nothing is captured when the shortcut isn't held.
Are text-rewrite commands ready to sell?
They work and they're useful, but they're an expanding group. We don't present them as final until quality settles.
Can I add custom commands?
Not in this beta wave. Custom commands are on the backlog; today the set is fixed to keep behaviour predictable.