Apple Dictation vs Akova for coding prompts
If you only need words typed now and then, Apple Dictation is free, already on your Mac, and good enough. If you dictate prompts into Claude Code, Cursor or Codex all day, you will hit three limits quickly: it stops when you pause to think, it does not know your project’s names, and it types your ramble exactly as you said it. Akova is built around those three limits.
Below is what each one does, checked against Apple’s own documentation, and a simple way to decide.
Checked on 12 October 2026 against Apple’s macOS Tahoe user guide and feature availability page.
What Apple Dictation does well
Start with the honest part: built in Dictation is better than many people think, and some comparison pages online repeat limits that no longer apply.
- It is free and already installed. Turn it on in System Settings, then Keyboard, then Dictation. Start it with the Microphone key, a shortcut such as pressing Fn twice, or Edit, then Start Dictation.
- It has no length limit. Apple’s guide says you can dictate “text of any length without a timeout.” Pages that quote a 30 second cap on total length describe older behaviour.
- It can run on your Mac. Apple’s guide says Keyboard settings show whether your dictation is “processed on your device and not sent to Siri servers.” On device dictation “requires download of speech models” and covers a shorter language list than standard Dictation.
- It punctuates for you. In supported languages it inserts commas, periods and question marks, and “new line” presses Return.
- It types into Terminal. That means it works with Claude Code and Codex CLI, which run in your terminal.
For a quick message or a one line commit description, that is all you need.
Where Apple Dictation gets in the way of a prompt
A prompt for a coding agent is not a text message. It is thirty seconds to two minutes of half formed thinking, full of file names, with pauses while you look at the code. That is where the built in tool starts to cost you time.
It stops when you stop. Apple’s guide says Dictation “automatically stops when no speech is detected for 30 seconds.” A long look at a stack trace ends your session, and you start again mid thought.
It does not know your repo. Standard Dictation has no list of your own
words. Names like useSessionStore, PostHog or stripe-webhook come out as
whatever the speech model thinks you said, and you fix them by hand in every
prompt. Voice Control has a vocabulary list, but Apple notes that standard
Dictation is not available while Voice Control is on.
It types exactly what you said. Every “um, no wait, actually” lands in the prompt. The agent reads all of it, and you either clean it up by hand or send a messier prompt than you meant to.
Your language may not be on device. The on device list covers English, French, German, Spanish, Japanese, Korean, Mandarin and a few dozen more variants, but not every language standard Dictation supports. Greek, for example, is not on it as of today. If your language is missing, check the line in Keyboard settings before you dictate anything private.
What Akova changes for coding agents
Akova, an on-device dictation app for Mac developers works the same way at the surface: you talk, the text lands at your cursor, in any terminal, editor or chat box. What differs is what happens between your voice and the cursor.
- Hold to talk. You hold a hotkey while you speak and release when you are done. A long pause to think does not end anything, because the recording lasts as long as you hold the key.
- Custom vocabulary, in the free tier. Add your repo names, APIs and jargon once, and they come out spelled right.
- Prompt mode. It turns a spoken ramble into a structured prompt before it pastes. On macOS 26 it uses Apple Intelligence. On macOS 14 and 15 it runs on a local Ollama model instead.
- On-device transcription. Akova transcribes with WhisperKit, which runs OpenAI Whisper models through Core ML on Apple Silicon. Your audio stays on your Mac by default, and dictation works offline.
- Its own language list. The picker offers 16 languages plus automatic detection, Greek included, all transcribed on your Mac.
The custom vocabulary list looks like this, one word per line:
akova-landing
useSessionStore
stripe-webhook
PostHog
WhisperKitAkova needs macOS 14 Sonoma or later on an Apple Silicon Mac (M1 and later). Intel Macs are not supported. Apple Dictation runs on more Macs, which is a fair point in its favour.
Price and setup, side by side
| Apple Dictation | Akova | |
|---|---|---|
| Price | Part of macOS | Free tier, or €40 once for Lifetime |
| Runs on device | Yes, for languages on Apple’s on device list | Yes, by default |
| Ends on a pause | After 30 seconds of silence | No, it records while you hold the hotkey |
| Your own words | No list in standard Dictation | Custom vocabulary, free |
| Ramble to prompt | No | Prompt mode, Lifetime |
| Works in Terminal | Yes | Yes |
| Requirements | Any Mac that runs macOS | macOS 14 or later, Apple Silicon |
The Akova Lifetime licence is a one time purchase with all future updates, activates on up to 3 Macs, and has a 14 day money back guarantee. Every Pro feature, Prompt mode included, is free for 7 days from your first launch, with no card and no account. After the trial, Akova keeps working on the free tier. The full list is on the Free and Lifetime pricing section.
Which one to pick
Stay with Apple Dictation if you dictate a few sentences a day, mostly outside your editor, and the words you use are ordinary English. It is free, and it will not get in your way.
Switch to Akova if you dictate prompts to a coding agent several times an hour, your prompts are full of project names, or you think out loud with long pauses. The vocabulary list and hold to talk fix the first two on the free tier. Prompt mode fixes the third.
If you are weighing other third party apps too, our Superwhisper alternatives roundup compares six of them, Apple Dictation included. For the setup inside your terminal, see how to dictate prompts into Claude Code on a Mac.
Try both on the same prompt
Pick the next task you were about to type into your agent. Dictate it once with Apple Dictation and once with Akova, then send whichever reads better. Download Akova free from the home page, add five words from your repo to custom vocabulary, and the comparison takes about five minutes.