| Current price | $9.99 once for one Mac, plus tax where applicable | Pro: $8.49 a month; $84.99 a year; $249.99 one-time lifetime licence |
|---|
| Pricing model | One-time purchase, with future Mac app updates included | A free plan, then monthly, annual or Lifetime. Its plans page says every paid plan has the same features and only the billing differs |
|---|
| Free plan | Yes: 2,000 words a week, then one free dictation a day until Monday. A 7-day trial of the Pro features on a new install | Yes: unlimited dictation with local Whisper models, up to two modes without AI processing, custom vocabulary and text replacements. New users get 3000 words of Pro features first |
|---|
| Lifetime or one-time option | The one-time purchase is the only paid plan, with no subscription | Yes: Lifetime at $249.99 once, and its plans page says Lifetime keys never expire. Whether Lifetime includes future major versions is not stated in public docs |
|---|
| Devices covered by one licence | One Mac per licence, which can be moved by deactivating it on the old Mac | One Pro licence covers unlimited Macs and Windows PCs, plus an iPhone and an Android phone |
|---|
| macOS | Yes (Apple silicon, macOS 13 or later) | Yes (macOS 14 or later) |
|---|
| Intel Mac | No: Apple silicon only | Not stated in public docs. Its pages say macOS 14 or later and do not mention Intel |
|---|
| Windows | No (planned, not available) | Yes (Windows 10 or 11, on x64 and ARM64) |
|---|
| iPhone and iOS | No (an iOS keyboard is in development) | Yes, as a keyboard. The minimum iOS version is not stated in public docs |
|---|
| Android | No (planned, not available) | Yes, on phones only, as a floating bubble. Tablets are not supported |
|---|
| Local speech recognition | Yes: on your Mac | Yes, with local models. Free includes the local Whisper models, and Pro adds further local models |
|---|
| Cloud model option | No cloud speech models. An optional advanced setting can send a transcript to an AI provider you configure, and it is off by default | Yes, on Pro: cloud voice and language models, or your own API keys |
|---|
| Offline | Yes, once its models are downloaded. A paid licence key is activated online once | Yes with local voice models. AI formatting offline needs a local language model in the Message or Mail mode, and other AI processing needs the internet |
|---|
| Account required | No account. A paid licence key is activated online once | Pro uses a licence key emailed after purchase and entered once per device. Whether the Free plan needs an account is not stated in public docs |
|---|
| Custom vocabulary | Yes: a Dictionary with 10 entries on the free plan and unlimited with Pro, plus "when I say, write" entries | Yes: custom vocabulary and text replacements, both on the free plan |
|---|
| Per-app behaviour | Per-app styles on Pro and the trial. In terminals and code editors only spoken punctuation and commands are applied | Modes can activate automatically by app, and each mode sets its own voice model, AI processing and prompt |
|---|
| Formatting | Corrections, filler-word removal, spoken punctuation, lists and numbers, with a polished mode and tones on Pro | Plain transcription on the free plan. AI-formatted modes such as Message, Mail and Note need Pro |
|---|
| Voice editing | Edit by voice on Pro and the trial: select text and say what to do with it, and Cmd+Z undoes the result | Super mode on Pro takes spoken commands on selected text for formatting and transformation. Its page says rewriting and summarizing do not work well there |
|---|
| Coding integrations | None. It types at the cursor, and in terminals and code editors it applies spoken casing and symbol commands | Claude Code, OpenCode, Codex, Cursor, Pi, Grok and DeepSeek on macOS, plus a command-line tool and MCP server on macOS |
|---|
| Meeting recording | No. It transcribes dropped audio files, up to 10 minutes on the free plan | Yes: a Meeting mode, system audio recording and optional speaker separation on Pro, plus file transcription on Pro |
|---|
| Model choice | No user-facing model picker. The speech model follows the language you choose, and Pro adds one optional, more accurate formatting model | Yes: a voice model and a language model per mode, local or cloud, in a models library, plus your own API keys on Pro |
|---|
| Languages | 25 European languages, plus Hindi | 100+ languages, per its introduction page |
|---|
| Main strength | A simple, local-first Mac dictation workflow at a low one-time price | The broader, more configurable platform: models, modes, coding agents, meetings and four platforms |
|---|
| Main trade-off | Mac only today, Apple silicon, one Mac per licence and few settings | More to configure, a higher price for Pro and Lifetime, and AI features that need Pro |
|---|