Speech to text for ADHD

How talking instead of typing can help

When you type, you are doing several jobs at the same time. You are holding the idea in your head, choosing the words, spelling them, watching for typos, and steering the sentence to a full stop. For some people that juggling is fine. For a lot of people with ADHD, it is exactly the load that makes writing exhausting, because every one of those jobs wants attention and the idea you started with quietly slips out the back.

Talking removes most of that load. You think and you speak, and the spelling and the typing happen later, separately. Writers and educators who work with ADHD, including resources like ADDitude and Understood, describe this as separating idea-generation from the mechanics of writing — your attention does one thing at a time instead of all of them at once. Many people with ADHD report that this is the difference between starting and not starting.

I want to be careful here, because the internet is full of tools promising to fix things they cannot fix. Speech to text does not treat ADHD. It will not help you focus, finish your taxes, or remember the dentist. What it does is narrow. It takes the specific moment where typing-the-first-draft is the wall, and it lowers the wall. If your wall is somewhere else, this is the wrong tool, and that is a fine thing to know early.

Getting the first messy draft out

The trick that makes dictation work for a brain that resists starting is permission to be bad. You are not writing the final email. You are talking out a rough version that you will clean up after.

So you ramble. You say "okay so the thing about Tuesday is, no wait, start with the budget, the budget is the part they care about." You back up, you repeat yourself, you trail off. That is fine. The goal is to get the shape of the thing onto the screen, because editing a messy paragraph is a completely different mental task than producing a blank one — and for most people it is the easier task by a wide margin.

This is also where dictation quietly beats a sticky note or a voice memo. A voice memo just moves the problem; now you have a recording you have to listen back to and transcribe. Speech to text gives you editable words immediately. You talk, the text is there, and you are already past the part that was stopping you.

The fastest way: a system-wide hotkey

Here is where a dedicated app changes the experience. Whisper by Remskill is not a browser extension or a single-app feature. It is a desktop app that works like a keyboard: press a hotkey, speak, and the transcript is pasted at the cursor, in whatever app you are in — your email, a document, a chat box, the note where you dump ideas.

That "any app" part matters more than it sounds for an ADHD workflow. You do not have to open a special dictation window, find the right mode, and lose your thread on the way there. The thought arrives, you hold a key, you say it, it lands where you were already looking. Fewer steps between thinking and done is the whole point.

Setup is short:

That is the whole loop. The transcript appears where your cursor is, you read it, you fix it.

Speak, and the words appear

Once it is running, the experience is plain in the best way. You put your cursor in the document, hold the key, talk, let go. A second or so later the words are sitting there, as if you had typed them, minus the part where typing them was the problem.

Because the local transcription runs on your machine — pure-Rust, no Python sidecar, no server in the loop — it works offline. There is no upload, no waiting on a connection, no spinner between you and the text. For a brain that loses the thread when a tool makes it wait, that immediacy is not a nice-to-have. The faster the words show up, the more likely the thought survives the trip.

Letting AI tidy the rambly draft

Spoken-out drafts are messy, and that is the point — but you usually do not want to send the mess. Whisper has an optional AI cleanup step that trims the filler and tidies the phrasing before the text lands. So "um, yeah, so the, the report is basically done, I think, we just need the numbers" becomes "The report is basically done; we just need the numbers."

For an ADHD draft, this is the part that closes the loop. You get to ramble freely, knowing the cleanup will catch the false starts, so you are never editing and producing at the same time. The cleanup runs locally through Ollama, free, on your own machine; Pro users can route it through the cloud instead, but the filler-trimming benefit is right there in the free local pipeline. You can also turn it off and paste the raw transcript when you want your exact words.

Keeping it private and offline

Some of what you dictate is not for anyone else. A first draft of a hard email. Notes about your own health. The thing you are only writing to get it out of your head. None of that should end up in a vendor's logs because you needed to type with your voice.

In Whisper's local mode, the audio and the transcription never leave your laptop. There is no cloud round-trip and no account-linked recording sitting on a server somewhere. Your laptop already has a microphone and a processor; for a paragraph of dictation, it does not need a server in the loop. If privacy is part of why typing-it-out feels heavy, local-first is the version that respects that.

What the full Whisper app looks like

The hotkey is the part you will use most, but there is a settings surface behind it. You choose a transcription engine — Whisper's models, whose multilingual variants cover 99 languages and whose English-only builds cover exactly one, or NVIDIA's Parakeet, about 600 MB and 5 to 10 times faster than Whisper on CPU, covering English plus 24 European languages. You set your hotkey, keep a history of what you dictated, and save presets. None of that is required to talk out one paragraph; it is there for when you want to tune it.

Free built-ins worth trying first

Before you install anything, your computer already does a version of this for free, and for a lot of people it is enough. On Windows, press the Windows logo key + H and voice typing opens in any text box, no subscription. On a Mac, the built-in Dictation shortcut does the same, and on Apple silicon it runs on-device. In any Google Doc, Tools then Voice typing turns on dictation right inside the document.

If you are dictating the occasional message or a short note, start there. It is free, it is built in, and it works. The one tradeoff worth knowing: Windows' Win+H needs an internet connection, while macOS dictation and Whisper's local mode do not.

A dedicated app starts pulling ahead when you are dictating a lot, when you want the filler cleanup, or when you want it to work offline and paste into every app the same way. Pick the smallest tool that solves your problem — for a one-liner, that is the key you already have.

What Whisper costs

The local dictation pipeline — transcription and the AI cleanup over Ollama — is free for any signed-in user, with no card at signup. So getting your voice onto the page with Whisper costs nothing. Whisper Pro adds the cloud features (OpenAI transcription, cloud AI enhancement, voice web search) and carries a separate trial. The exact numbers live on the pricing page rather than here, because prices move and a blog post is a bad place to keep them current.

Speech to text will not reorganize your life. It does one small, specific thing: it lets you talk when typing is the wall. For some people that is the difference between a draft and a blank screen they have been avoiding since Tuesday. Try the free version your computer already has. If the wall keeps showing up, reach for the tool that knocks it down faster.