The question everyone asks

Apple Dictation is free and very capable. When is VoiceType worth it?

This page does not claim Apple is “bad.” It answers the #1 buyer objection: if Apple already ships dictation, why pay? The short answer: because professionals often need a tool tuned for terminal workflows, consistent on-device Whisper quality, and optional polish—with a product story that is only dictation, not a slice of a giant OS.

Honest comparison

Apple’s experience varies with macOS version, language, and whether you use on-device or server-based recognition. VoiceType is built to keep Whisper-class transcription on your Mac and get text to the cursor in demanding apps. Your mileage always depends on your hardware and your exact settings.

Dimension VoiceType Apple Dictation
Accuracy in noise / accents Uses on-device Whisper (WhisperKit). Strong general-purpose STT; you control model size vs speed in preferences. Often excellent on recent Macs; quality depends on language pack, on-device vs server, and app. Very good for many users out of the box.
Terminal, shells, secure fields Designed and documented for Terminal, iTerm2, and IDE terminals, including secure input fallbacks (clipboard where macOS blocks injection). Can work in some terminal contexts, but is not a “terminal product.” Secure and password fields may block or behave inconsistently—exactly the edge cases we optimize for in docs.
Code, commands, and symbols Whisper is strong at what you say; for code-heavy dictation, pair with your editor’s own tools. We focus on getting raw text in fast—see coding notes. Fine for plain text; symbol-heavy or mixed content can be fiddly compared to a workflow built around a dev’s muscle memory and clipboard recovery.
Hotkey and flow Global: double-tap , or optional +Space / +Space in settings—built for one consistent gesture everywhere. Typically fn-fn or keyboard dictation shortcut; can conflict with other OS features; less opinionated for “developer hotkey” parity across every app.
AI polish and formatting Optional AI Polish (Pro) for grammar and clarity; on-device Qwen 3.5 or other configured paths. Core dictation never requires the cloud. Apple Intelligence and writing tools in supported apps help with text in those apps—not the same as a single dedicated polish pipeline for every text field.
Offline quality Whisper models run on-device; offline use is a first-class story once models are downloaded. On supported Macs you can enable on-device dictation in Keyboard settings; not all features or languages are equal offline. Check your locale and “Dictation” / “Siri & Dictation” settings.
Where text lands “Paste at cursor” goal with accessibility-based insertion (direct build) and clipboard where required—see known limitations. Inline replacement in supported fields; behavior differs by app and field type (e.g. browser vs native).
History and re-use In-app history for recent phrases (so you can recover or reuse without re-dictating). No first-party “dictation history” product for arbitrary apps in the way a standalone dictation app can offer.
Privacy posture (high level) Core audio and transcription on your machine; we do not run a “log of everything you said” on our servers for the default path. Optional polish may use cloud or local models per your choice. Apple’s privacy program is strong; server-side dictation or analytics may still apply depending on settings and feature use. If you are evaluating pure “nothing leaves the machine,” review Apple’s own documentation for your exact configuration.

When Apple Dictation is enough

Under those conditions, Apple is an excellent free option—and we are not afraid to say it.

When you have outgrown it

Try VoiceType free. If Apple already wins for you, no hard feelings—uninstall and move on.