Best voice typing software: 7 tools ranked

How I picked these seven

Selection first, ranking second. To make the list, a tool had to be a real, shipping product that turns speech into text where you write. Not a category, not a discontinued app, not a thin wrapper that disappears next quarter. The browser toys and abandoned freeware that pad most vendor lists did not make the cut.

The ranking logic is boring on purpose:

1. Where it types. System-wide tools that work in every app rank above tools locked to one document editor.
2. What the free tier is worth. A free tier you can live in ranks above a free teaser with a word meter.
3. Offline behavior. Dictation that survives a dead Wi-Fi connection ranks above dictation that does not.
4. Platform coverage. Windows and Mac both, where possible.

No score out of ten, no weighted matrix pretending this is science. Four criteria, applied in order, ties broken by my own daily use (a sample size of one, which is still one more than most vendor lists). Your mileage will vary, and the final section is written for that case.

The seven tools, ranked

1. Whisper by Remskill: unlimited free local dictation, system-wide

Whisper is a desktop app for Windows and macOS. Press a hotkey, speak, release, and the text lands wherever your cursor is: email, Slack, your CRM, a code editor, anywhere. The default hotkey is Ctrl+Space on Windows; on a Mac you hold Command and Option together like a walkie-talkie and release to stop.

The reason it ranks first is the free tier. It is not a metered teaser. The entire local pipeline is free for signed-in users: all eight local Whisper models, the Parakeet engine, AI text cleanup through Ollama, history, custom vocabulary, the hotkey. Everything runs on your machine, nothing leaves your device, and it works with the internet off.

Multilingual Whisper models cover 99 languages — a figure that applies to the multilingual variants; the English-optimized builds do English only. If you speak English most of the day and want speed, the NVIDIA Parakeet engine (~600 MB) runs 5–10× faster than Whisper on CPU and covers English plus 24 European languages. The two engines make different trade-offs; the Whisper vs Parakeet comparison goes deeper. A paid Pro tier adds OpenAI cloud transcription and web search with your own API key; details live on the pricing page.

Now the honest part. Whisper is desktop-only, with no iPhone or Android app. On Mac it requires Apple Silicon; Intel Macs are out, and no Linux build exists. If you dictate from your phone more than your laptop, or your laptop is a 2017 MacBook, skip Whisper and read on. I dictated that sentence with Whisper, which made it sting twice.

The hotkey loop, the engine lineup, and what stays on your machine are all laid out on the voice typing page for Whisper.

2. Wispr Flow: the polished cross-device option

Wispr Flow is the tool I would point you to if you live across devices. It runs on Mac, Windows, iPhone, and Android, auto-detects between 100+ languages mid-flow, and markets itself on turning messy speech into polished text at 4× typing speed. The phone apps are the real differentiator. Nothing else on this list follows you from a desktop to a lock screen.

The catch is the meter. The free tier caps at 2,000 words per week on desktop, which an average dictation user burns through in a day or two. Past that it is a Pro subscription at $15 per user per month, or $12 on the annual plan. It is also cloud-based: Wispr's own privacy page says transcription always happens in the cloud, so your audio leaves your machine. If phone dictation matters more to you than offline or unlimited-free, this is your pick. Skip Whisper without guilt.

3. Apple Dictation: the built-in Mac answer

Already on every Mac. You start it from the Microphone key, a customizable keyboard shortcut, or Edit > Start Dictation, and on current macOS you can dictate text of any length without a timeout; it stops itself after 30 seconds of silence.

Apple's dictation guide shows where to check whether your dictation is processed on-device instead of being sent to Apple's servers.

For a 40-word reply in Messages, this is the right tool, and paying anyone for that use case is silly. The limits show up at the edges: availability varies by language and region, Apple's guide documents no custom vocabulary for your own product names, and the whole thing lives inside the Apple fence. If your dictation is short, occasional, and on a Mac — stop here and keep your money.

4. Windows voice typing: press Win+H and talk

Microsoft's built-in voice typing starts with Win+H in any text box, ships inside Windows 10 and 11 with nothing to download, and inserts punctuation automatically as you speak. It covers more than 40 languages. For a free feature most Windows users have never pressed, it is better than its reputation.

Two caveats. It requires an internet connection, as Microsoft's own support page says, which means it dies the moment your train Wi-Fi does. And Microsoft documents no custom vocabulary for it, so expect it to keep mishearing your company's product name. Fine for occasional use; frustrating as a daily driver.

5. Google Docs voice typing: free, but fenced into Docs

Open a Google Doc, click Tools, then Voice typing, and a microphone box appears. It works in current Chrome, Edge, and Safari, handles a long published list of languages and accents, and covers speaker notes and captions in Google Slides. For students and anyone who drafts everything in Docs anyway, it costs nothing and works today.

The fence is the whole story. It types into Google Docs and Slides, not into your email client, not into Slack, not anywhere else. Your browser controls the speech-to-text service, so quality follows the browser, not you. The boring truth is that a tool you cannot use in 90% of your typing is a feature, not a voice typing setup.

6. Superwhisper: the indie local option for Apple-first users

Superwhisper deserves more attention than it gets. It runs on macOS, Windows, and iOS, works offline, and activates with Option+Space or push-to-talk in any app. The free tier includes voice-to-text in any app, meeting recording, 100+ languages and dialects with translate-to-English, and unlimited use of its small AI models. Larger models and custom API keys sit behind paid Pro and lifetime plans.

It overlaps with Whisper at #1 more than anything else here: both are local-first, hotkey-driven, with free tiers you can live in. The differences sit at the edges. Superwhisper grew up Mac-first with iOS in the family, while Whisper's free tier includes its full local model lineup on both desktop platforms. If you are deep in the Apple ecosystem and want an iPhone companion, Superwhisper is a fine choice over us.

7. Dragon: the professional license that started it all

Dragon Professional v16 is the descendant of the software my relative shouted at in 1998, and to Nuance's credit it stopped asking for 45-minute training sessions long ago. Nuance claims you can dictate documents 3 times faster than typing with up to 99% recognition accuracy, and the product is optimized for Windows 11, backwards-compatible to Windows 10. No Mac version appears on the product page, and pricing is a professional license arranged through their sales channel rather than a public price tag.

Here is my one strong take for this article: if a dictation tool needs you to train it, correct it, or enroll your voice profile, it is solving a 1999 problem. Modern speech models work out of the box on most accents — Whisper-family models ship with 99 languages and zero enrollment. Dragon in 2026 earns its place for professionals with established Dragon workflows, legal and medical teams with years of custom commands, not for someone starting from zero.

Quick comparison

One row deserves a footnote: "free tier" means what you can keep using indefinitely, not what a trial lends you for two weeks.

Why system-wide beats in-app dictation

The biggest split in this list is not accuracy. It is territory. Google Docs voice typing is locked to two Google apps. Apple Dictation stays inside the Mac. The system-wide tools — Whisper, Wispr Flow, Superwhisper, Windows voice typing — type wherever your cursor blinks.

That territory difference decides whether voice typing becomes a habit. The average dictation user speaks around 145 words per minute against roughly 40 typed, but the speedup compounds when the tool works in every box you type in: the email, the ticket, the chat reply, the commit message. A tool you must switch documents to use adds the friction it was supposed to remove. Most productivity tools are typing problems in disguise; a system-wide hotkey is the rare fix that removes the typing instead of decorating it.

How I'd pick, case by case

Strip away the ranking and four questions remain. Dictating from your phone most of the time? Wispr Flow — nothing else here does lock-screen-to-laptop. Short occasional notes on one platform? Use what came installed: Apple Dictation on a Mac, Win+H on Windows, Tools > Voice typing in Docs, and pay nobody. Long-form daily dictation where privacy or offline matters? That is the local-first pair: Whisper if you want the full free model lineup on Windows and Mac, Superwhisper if you want an iOS companion in the Apple world. An established Dragon workflow with custom legal or medical commands? Stay on Dragon; migrating years of muscle memory costs more than the license.

Still unsure? Start free. Every path above has a no-cost entry, and the free dictation software guide walks through them in detail. Upgrade when you hit a real wall, not a marketing one.

Seven tools, but only three decisions: built-in or dedicated, cloud or local, desktop or everywhere. Answer those and the list collapses to one name on its own. My relative needed 45 minutes of training and a flying headset to learn that dictation was not ready in 1998. You need about ten minutes and zero dollars to learn that it is now — which, as product timelines go, is at least the right direction.