Learn
One concept at a time.
Reference-grade explainers on the terms that come up around dictation on a Mac. Each entry defines the term and shows where it lands in the day-to-day.
-
Accessibility-API typing
Accessibility-API typing, defined
The macOS mechanism a dictation tool uses to type into whichever text field your cursor is in, without copying through the clipboard.
-
BYOK cloud transcription
Bring-your-own-key cloud transcription
A cloud transcription mode where the user supplies their own API key for an AI provider, so the request goes from the user's Mac to the provider they pay, not through the dictation tool's servers.
-
Custom dictation dictionary
Custom dictation dictionary, defined
A user-maintained list of replacements that runs on every transcription to fix the words a general-purpose speech model gets wrong.
-
Dictation for RSI
Dictation for RSI and wrist health
Using dictation to shift typing load off your hands when repetitive-strain injury or general wrist fatigue makes a keyboard-only day untenable.
-
Dictation vs transcription
Dictation vs transcription
Dictation types your voice into the field you are already in; transcription produces a document from a recording you made earlier. Different jobs, different tools.
-
Mac dictation privacy
Mac dictation privacy, side by side
Where each Mac dictation option sends your audio, what it stores, and what you can do about it.
-
On-device dictation
On-device dictation, defined
Dictation that turns your voice into text on your own Mac instead of routing the audio through a cloud server.
-
Push-to-talk hotkey
Push-to-talk hotkey, defined
A keyboard combination you hold to record and release to transcribe — the input shape that makes a dictation tool feel like a part of the keyboard.
-
Voice coding
Voice coding, defined
Using voice as a second input alongside the keyboard while writing software — for the prose around the code, the AI prompts, and the commit messages.
-
Whisper family of speech models
Whisper models, explained
The Whisper family is a set of open speech-recognition models from OpenAI, released in several sizes that trade speed for accuracy.