Skip to content

How dictate compares

A fair, factual look at the macOS dictation landscape as of 2026. If anything here is wrong or out of date, please open an issue or send a PR — this page is meant to help people pick the right tool, not to sell anything.

At a glance

dictate Apple Dictation MacWhisper Superwhisper Wispr Flow
Licence MIT, open source Proprietary Proprietary Proprietary Proprietary
Price Free Free (bundled) Free + paid tiers Paid (subscription) Paid (subscription)
Runs offline ✅ default ⚠️ "Enhanced" only ✅ ✅ ❌ cloud
Telemetry None Per Apple's privacy policy None claimed None claimed Cloud by design
ASR engine Whisper (faster-whisper) Apple SFSpeech Whisper Whisper + others Proprietary
LLM cleanup Optional, local (Ollama) or BYO key (OpenRouter) None Optional, paid Built-in Built-in
Hotkey modes Hold / tap / double-tap Tap (fn) Configurable Configurable Configurable
Per-app vocab / presets ✅ ❌ Limited ✅ ✅
Voice commands in-utterance ✅ ❌ Limited ✅ ✅
Secret redaction before LLM ✅ n/a ❌ ❌ ❌
Auditable code ✅ ❌ ❌ ❌ ❌
Packaged .app ⏳ planned ✅ (built-in) ✅ ✅ ✅

When to pick each one

Pick Apple Dictation if the built-in tool already does what you need. It's free, it's there, and "Enhanced Dictation" runs locally on Apple Silicon. For most casual users, this is the right answer.

Pick MacWhisper if you want a polished, native Mac app for files (meeting recordings, podcasts, voice memos) and only occasional live dictation. It's the strongest tool in that category.

Pick Superwhisper or Wispr Flow if you want a managed product with support and you don't mind paying a subscription. Wispr Flow is the slickest hotkey-dictation experience on the market right now — if your threat model is fine with cloud transcription, it's hard to beat on UX.

Pick dictate if any of these matter to you:

  • You want the LLM cleanup pipeline disabled by default and full control over which model touches your text.
  • You want to audit the code that records your microphone, sends it to a model, and pastes the result. Every line is on GitHub.
  • You want secret redaction (API keys, AWS keys, tokens) to happen before anything optional leaves your machine.
  • You're a developer who wants per-app vocab presets, code-grammar mode, voice commands, and a scriptable URL scheme / Raycast / AppleScript surface.
  • You're fine running it from source (or building your own .app).

What dictate is honestly not great at (yet)

  • No signed .app. Until a paid Apple Developer ID is in the picture, install is git clone + ./install.sh. Friendly for developers, friction for everyone else.
  • No long-form file transcription UX. dictate is built for short, hotkey-triggered insertion. For "transcribe this 90-minute meeting", MacWhisper is a better tool.
  • No iOS / iPad story. macOS only by design.
  • Setup involves macOS permissions (Accessibility, Input Monitoring, Microphone). Required for any tool that types into other apps; dictate has a permissions guide and dictate doctor to help.

Migration notes

Coming from Wispr Flow or Superwhisper? The closest equivalent to their "AI rewriter" is dictate's optional cleanup step. It's off by default; enable it in the WebUI and pick a local Ollama model (e.g. llama3.2:3b) or wire in an OpenRouter key. Per-app presets in dictate correspond loosely to "modes" in those products — configure them via per-app vocab.