Everyone has had the experience of thinking faster than they can type. You are mid-thought, fingers flying, and the idea outpaces your keyboard. An app that writes what you say solves this problem at its root: instead of translating thoughts into keystrokes, you just speak, and the words appear. The question is which app actually delivers on that promise reliably enough to make it a core part of your workflow.
There are many voice-to-text tools on the market, and they differ enormously in accuracy, speed, and practical usability. This guide covers what to look for, what the major options actually do, and why the approach matters as much as the technology.
What Makes a Good App That Writes What You Say
The most important quality is accuracy. An app that makes you correct every third word is not saving you time — it is just shifting the work from typing to editing. Accuracy depends on the underlying speech recognition technology, the quality of audio processing, and how well the system handles different accents, speaking paces, and ambient noise.
The second quality is speed. There is a meaningful difference between an app that transcribes in real time as you speak and one that makes you wait two or three seconds after you finish a sentence. Real-time transcription feels like an extension of your voice. Delayed transcription feels like using a service.
Third is integration. The best voice-to-text app is the one that works everywhere: in your email client, your word processor, your browser, your notes app, your terminal. App-specific solutions require you to switch contexts to use them. System-level solutions work wherever your cursor is.
Apple's Built-In Dictation
Every Mac and iPhone includes built-in dictation, and it has improved significantly over the past several years. On newer Apple Silicon Macs, on-device processing means your voice never leaves the device. The accuracy is decent for everyday prose, and the latency is low.
However, built-in dictation has real limitations. It struggles with technical vocabulary, proper nouns, and specialized terminology. The activation model — double-pressing a key or the Globe key — can feel clunky when you want to rapidly switch between speaking and typing. And on older hardware, quality drops noticeably.
Browser-Based Voice Typing
Several web services offer voice-to-text input directly in the browser. These are convenient for one-off transcription tasks but are not practical for general-purpose dictation. You cannot use them in native apps, you need an active browser tab, and they typically impose session limits. They are fine for transcribing an occasional voice memo, but they do not replace a proper dictation tool for daily use.
Why Hold-to-Speak Changes Everything
Most voice-to-text apps use a toggle model: press once to start, press again to stop. This creates a problem when you are typing and speaking in alternation, which is how most real writing sessions actually work. You type a heading, then speak a paragraph, then type a correction, then speak another paragraph. With a toggle model, you are constantly managing the on/off state, which requires attention that should be going to your content.
The hold-to-speak model is fundamentally better for mixed typing and dictation workflows. You hold a hotkey and speak. The moment you release, transcription stops. The behavior is instant, intuitive, and requires zero mental overhead. It is the same pattern as a walkie-talkie or a push-to-talk intercom, and it works for exactly the same reason: you always know whether you are transmitting because you can feel the key under your finger.
Steno: A System-Level App That Writes What You Say
Steno is a Mac and iPhone app built around this hold-to-speak philosophy. It lives in your menu bar on Mac and works as a keyboard extension on iPhone. Hold the hotkey, speak, release — your words appear wherever your cursor is, in any application. There is no mode switching, no browser tab, no dedicated window to manage.
The accuracy comes from processing that handles real-world speech conditions well: accents, filler words, background noise, and rapid speech. Steno also includes a Smart Rewrite feature that can clean up dictation automatically — removing false starts, correcting capitalization, and applying context-appropriate formatting — before inserting text into your document.
On iPhone, Steno works as a custom keyboard, so the same hold-to-speak experience is available in Messages, Notes, Mail, and any other app that accepts text input. This is more practical than the native iOS dictation microphone button, which is small, hard to hit, and toggles rather than holds.
Who Benefits Most
An app that writes what you say is useful for anyone who types a lot, but some groups see especially dramatic improvements.
Writers and Content Creators
People who produce large volumes of text — bloggers, authors, journalists, screenwriters — find that dictation removes creative friction. Ideas flow more naturally in speech than in typing. A first draft spoken aloud tends to have more energy and authenticity than one typed character by character.
Professionals with Repetitive Documentation
Doctors, lawyers, and consultants who write similar documents repeatedly find dictation dramatically reduces the time spent on documentation. When you know the structure and just need to fill in the details, speaking is always faster than typing.
People with Typing Fatigue or RSI
Repetitive strain injuries and conditions like carpal tunnel syndrome make typing painful. Voice-to-text is not just a convenience in these cases — it is a way to continue working without aggravating an injury. An app that writes what you say can mean the difference between a productive workday and one spent in discomfort.
Students and Researchers
Taking notes, summarizing readings, and drafting papers all benefit from the speed advantage of speech. Students who dictate their notes and first drafts consistently report that they write more and edit less than when they type.
Getting Started with Voice-to-Text
The biggest obstacle to adopting voice-to-text is psychological rather than technical. Most people have a mental model of dictation as slow and error-prone based on early experiences with inferior tools. Modern voice recognition is dramatically better, and a few hours of practice is enough to make the habit feel natural.
Start with low-stakes writing: notes to yourself, quick emails, to-do lists. Once you are comfortable speaking naturally into the microphone without feeling self-conscious, move to longer-form writing. Within a week, most people find they are speaking at a rate equivalent to 100 words per minute or more — roughly three times the average typing speed.
The fastest path from thought to text is not a better keyboard — it is eliminating the keyboard from the equation entirely.
If you are ready to try an app that writes what you say, Steno is available at stenofast.com. Download it, set a hotkey, and say your first sentence. The experience speaks for itself.