This Free Mac App Combines Several AI Transcription Tools

If you've ever wished the text-to-speech feature on your Mac worked better, Spokenly might be the solution for you. It's a free Mac application that lets you use modern AI transcription right on your Mac for free—just trigger the application with a keyboard shortcut, say what you want to type, and watch as it appears on the screen.

Whatever you feel about modern AI, it's hard to deny that it's good at turning the spoken word into text. Whether it's real-time transcription or creating a transcript for a video or audio file, the open source Whisper engine is accurate and fast. Spokenly is a free application that lets you use this technology on your Mac.

Install the application, which is currently free on the Mac App Store, and it will start in your menu bar. Transcribing is simple: Just press and hold the right Command button and you'll see the pop-up, as shown above. Say what you want to type and let go when you're done—the text will appear in the currently active text field.

You could honestly just install the application, use it this way, and never think about the settings—it mostly just works. If you want to dig into, though, open the settings by clicking the menu bar icon and clicking General Settings. You can decide whether the application will show up in the menu bar, dock, or both. You can decide which microphone you'd like the application to favor, assuming you use more than one. And you can decide whether transcribed text should be copied to your clipboard by default.

If you like experimenting with AI, though, you should check out the Voice Models tab. From here, you you can choose which AI model you want to use. By default, the application uses an online version of "Whisper Large v3," which is the only model that supports showing your transcription in the pop-up window as you're talking. There are other models you can use, though, including GTP-4o via Spokenly.

Credit: Justin Pot

If you'd rather not use an online service, there's support for downloading and running various Whisper models locally. You will have to find a balance between hard drive space, performance, and accuracy—I recommend trying out a few models until one works for you. There's also support for using Deepgram, Fireworks, and OpenAI API keys, assuming you're already paying for one of those services.

If you don't know what any of that means, don't worry: The default model works extremely well. And there's more to dig into with this application. The Transcribe File tab of the settings window lets you add any video or audio file and get a transcription, with or without timestamps. There's even support for generating subtitle files for your videos.

Credit: Justin Pot

And there's a few power-user tools you can set up. The Quick Commands tab lets you choose a few combinations of words that run an action instead of transcribing. Just choose a trigger phrase and what you'd like that phrase to do—open a URL, launch an application, or run one of your Apple Shortcuts. For example, you could set up "Open Lifehacker" as a trigger phrase and our homepage as the URL. Do that and the Lifehacker website will open every time your say the words "Open Lifehacker," improving your life (one assumes).

Credit: Justin Pot

Another feature, AI Prompts, allows you to set a prompt that will change your text after dictating it. Just choose a separate keyboard shortcut and type your prompt for the text. For example, you could write a prompt that says "translate this text into Spanish"—you can now use a custom keyboard shortcut to transcribe something before using AI to translate it. Any prompt you can think of will work.

All of these features reward tinkering, which is exactly the kind of application I like. And the application is completely free as of this writing, while offering features that allow you to use it privately. Try it out if you wish you could get more out of transcription software.

View the full article