Voice to text for arthritis

Why people reach for voice to type less

I'll keep saying the disclaimer because it matters: this is not medical advice, and dictation software heals nothing. What it does is reduce the number of keys you press in a day. You speak, the computer types, and the keystrokes you would have made, you skip. For an inbox you'd normally answer over forty minutes of typing, that's a few hundred presses you simply don't make. That's the entire, boring benefit, and it's the honest one.

The job people actually want done is bigger than email. It's the long document you keep putting off because your hands aren't up for it. It's the chat reply, the form, the note you'd jot if jotting didn't cost anything. It's capturing an idea before it's gone, at the speed you think it, instead of the speed you can type it. When the keyboard is the uncomfortable part, handing the text to your voice is a way to keep working without it being a chore.

There's a speed side effect worth one sentence. Speaking runs around 145 words a minute for most people; typing is closer to 40. So beyond resting your hands, you tend to move roughly three and a half times faster, which is a pleasant bonus when the slower option was also the one that bothered you. None of that is a health claim. It's just arithmetic about keystrokes.

Press a hotkey, talk, the text lands in any app

This is the whole mechanic, and it's deliberately boring. You press a hotkey, you speak, you release, and the transcript pastes at your cursor, in whatever text field has focus. Whisper holds a short tail after you let go of the key, so your last word doesn't get clipped. Because it pastes at the OS cursor, every app is just "any text box" — your email client, a Word document, Slack, a browser form, your notes app. Same key, same flow, everywhere.

That's the part the landing pages overcomplicate. There's no plugin to wedge into each program, no API token to paste, no separate window to fish your words out of. Your cursor is in the box, you talk, the words appear in the box. A small capsule shows up while you speak so you know it's listening:

The hotkey is the one thing worth getting right up front, and it's also where comfort comes in. On Windows the default is Ctrl+Space; on Mac it's Command+Option, a modifier-only push-to-talk you hold while speaking. If holding a chord down is itself uncomfortable, you don't have to — switch it to tap-to-toggle in Settings, Recording, so one tap starts and one tap stops, and you never hold a key at all. (Every hotkey is customisable because I shipped a hardcoded one first and it collided with someone's music software at two in the morning. I have a master's degree.) If you've set up dictation on Windows or on Mac before, this is the same muscle memory pointed everywhere at once.

Set it up in two minutes (Windows or Mac)

You need a Mac on Apple Silicon or a Windows 10-or-newer PC, a working microphone, and whatever app you want to type into open in the background. The whole local pipeline is free for any signed-in account, with no payment method asked for at sign-up. Here's the sequence.

The slow part is the model download, not the setup. Everything else is the four steps above. Once it's running, the act of getting a thought onto the screen stops being a typing task and becomes a talking task — which is the whole point when typing is the uncomfortable bit.

voice to text on Windows · on Mac

One hotkey across every app you already use

The reason a system-wide tool beats a per-app feature is that it doesn't make you relearn anything when you switch programs. The same key that fills your email compose box fills a Word document, a Slack message, a browser form, a spreadsheet cell, and a commit message. As far as your computer is concerned, you're typing — so it works wherever typing works. One tool, every text field, on both Windows and Mac.

That matters more than it sounds when the goal is to rest your hands. If each app had its own dictation button, you'd be hunting for a different control all day, and half of them wouldn't exist. With one hotkey, the friction of starting drops to near zero: tap, talk, done. The fewer steps between "I want to write this" and "it's written," the less you reach for the keyboard out of habit when your hands would rather you didn't.

The honest scope, so there's no surprise: this puts words where your cursor is. It does not move the cursor, click menus, or navigate windows for you. For most people the bulk of the keyboard load is the text itself — emails, docs, messages, notes — so handing off the text already takes most of the pressure off. If you need the computer to do more than that by voice, there's a section below that points you somewhere better.

Local or cloud: which mode to pick

Start with local mode. A lot of what you'll dictate is personal — a note to your doctor's office, an insurance form, a message to family — and there's no reason that should leave your laptop to become typed text. Local transcription runs entirely on your machine, offline, with nothing sent to a server. If your Mac is Apple Silicon or your PC is from the last few years, local handles everyday dictation without complaint, and cloud becomes the escape hatch rather than the default.

Here's how the three paths differ, because the app makes you pick and I'd rather you pick well:

• Local Parakeet — NVIDIA's TDT engine, around 600 MB, and the fastest local option — 5 to 10 times faster than Whisper on CPU. Covers English plus 24 other European languages, 25 in total. No translate-to-English. If you write in English or another European language, this is the quick, fully offline pick.
• Local Whisper — slower than Parakeet on the same machine, but the multilingual builds cover 99 languages and can translate to English. The English-only builds are English-only, not 99. Pick this for Chinese, Japanese, Korean, or any translation work, which Parakeet can't do. Default English model is around 480 MB.
• Cloud (OpenAI, BYOK) — best accuracy and web access, using your own OpenAI key billed straight by OpenAI. Transcription runs on gpt-4o-mini-transcribe by default. Needs internet, so it's the one path that leaves your machine. The Cloud surface is part of Whisper Pro.

The boring truth is that for most everyday text, local is plenty. Both local engines run fully on your machine with nothing sent anywhere, which is the right default when you're typing personal things by voice. Cloud earns its place when you want top-tier accuracy on a hard recording or you need the model to pull a fact off the web mid-sentence. For a day of email and notes, start local and only reach for cloud when local leaves you wanting.

AI cleanup so you're not fixing it by hand

This step matters more for resting your hands than it first looks. Spoken language is messy. You say "um," you restart sentences, you trail off. If you then have to go back and fix all that by typing, you've put the keystrokes right back — which defeats the point. So Whisper has an optional AI pass that trims filler and tidies phrasing before the text pastes. Fewer corrections means fewer keys.

Windows Voice Typing adds basic punctuation as you speak, and macOS Dictation handles it when you say "comma" or "period." For heavier cleanup — stripping the filler, fixing run-ons, turning a spoken paragraph into something you'd actually send — Whisper runs that AI pass on request. Say the activation phrase "Hey whisper" and the text gets enhanced before it lands. On a local model that runs through Ollama, free, on your own machine; in cloud mode it's gpt-5-mini by default.

The point of the cleanup, for this use, is that you read the result once and move on instead of going back into the text with the keyboard. You can also turn it off and paste the raw transcript when you don't care about polish. Either way, the editing you do by hand drops, which is the part your hands will notice.

That same speak-then-clean flow pays off across everything you write — you can also type faster with your voice so a long message becomes a few spoken sentences instead of a paragraph you press out key by key.

When to skip a dictation tool like this

Sometimes the right tool isn't mine, and pretending otherwise would be doing you a disservice. Two cases come up often, and dictation-into-text is the wrong answer for both.

First, if you want to control the whole computer by voice — move and click the mouse, open apps, navigate menus, scroll, not just put text in a box — a dictation tool won't get you there. Whisper handles the text; it does not drive the machine. For full hands-free use, look at your operating system's accessibility tools first: macOS Voice Control lets you control the entire interface by voice, and Windows has Voice Access. Beyond the built-ins, Dragon (Windows) adds voice commands for the mouse and menus, and Talon Voice (Mac, Windows, Linux) goes furthest, pairing voice commands with eye tracking and noise-based clicking for true hands-free control. If that's what you need, start there, not here.

Second, if you only need to dictate the occasional short message, don't install anything yet. On Windows, press the Windows logo key + H and the built-in voice typing bar opens wherever your cursor is; it punctuates on its own and is free, though it routes through Microsoft's servers and needs internet. On a Mac, Dictation lets you speak into any text field, set up in System Settings under Keyboard, and on Apple Silicon it can run on-device. And for quick phone use, your phone keyboard's microphone already dictates into any field. A dedicated tool earns its place at volume — long writing, the lowest-friction key, offline use, and one hotkey that behaves the same everywhere. Below that bar, use what's free.

The framing here is the same one I use in the broader guide to dictation software for RSI — reduce the keystrokes, keep it honest, and reach for a heavier tool only when the keyboard, not just the typing, is the problem.

None of this is a fix for arthritis, and I'm not going to pretend it is. It's a way to make text without pressing keys, which is a small, useful thing when pressing keys is the uncomfortable part. The cursor is the integration: talk into any box, get text, clean it up without going back to the keyboard. I dictated most of this guide into a text editor with the same hotkey, hands mostly off the keys, then read it back once. That's the whole trick — and if your joints need more than a typing break, the people to ask are the ones with the medical degrees, not me.