Dictation

Talk normally.
Get clean text.

Speaking runs roughly 3x faster than typing. Vox keeps every word — filler removed, formatting done, pasted right where your cursor already is.

Listening…

You say

"uh so, can you like, take a look at the, um, the deck before three"

→

Vox types

Can you take a look at the deck before 3?

Code-switch mid-sentence. Vox keeps up.

Auto language detection runs on every recording — no manual language picker to fumble with before you speak. Mixing English and Tagalog in the same sentence works exactly like speaking one language.

"Pwede ba you send that bago mag-meeting tomorrow?"

Any app. Wherever your cursor is.

Vox pastes straight into the field you're focused on — Mail, Slack, Notes, a terminal, a browser form. No copy-paste, no switching windows.

When something's off, you'll know.

If a backend call fails, Vox falls back automatically and shows a clear signal on the HUD instead of guessing silently. You always know whether what landed is what you actually said.

Vox HUD
Transcribed and pasted
Fell back to on-device model — backend unavailable

Press ⌃⇧Space
and start talking.

No account needed to start. Up to 2,000 words a week, forever.

Download for Mac