The ability to speak to type — to open a text field, say what you want, and watch accurate text appear — has been theoretically possible for decades. But it has only become genuinely fast and accurate enough to be a real workflow replacement in the last few years. Today, for most knowledge workers on a Mac, speaking is the fastest, lowest-friction way to input text. The bottleneck is no longer the technology. It is building the habit.
This guide covers everything you need to set up, practice, and internalize speak-to-type as your primary input method. By the end, you should have a clear path to tripling your effective writing speed without any additional effort.
Why Speaking Is Faster Than Typing
The numbers are straightforward. The average person speaks at 130 to 150 words per minute in natural conversation. The average knowledge worker types at 40 to 60 words per minute. Even the fastest touch typists rarely sustain more than 100 words per minute over extended periods. When you speak to type, you operate at your natural speaking pace — roughly two to three times faster than most people type.
Beyond raw speed, speaking engages a different cognitive mode than typing. When you type, your brain simultaneously manages physical key selection, spelling, sentence construction, and idea generation — all competing for the same limited working memory. When you speak, the physical mechanics disappear almost entirely, freeing cognitive resources for the ideas themselves. Many users report that dictated first drafts are not just faster but actually better structured and more fluent than typed ones, because the natural forward momentum of speech discourages the over-editing and second-guessing that typing encourages.
Setting Up Speak-to-Type on Mac
Hardware: Microphone Matters
The single biggest factor in speak-to-type accuracy, after the quality of the software, is microphone quality. The built-in microphone on a MacBook is surprisingly capable in a quiet room, but it picks up keyboard noise, fan noise, and ambient sound that degrade accuracy in real working conditions. A simple wired headset with a boom microphone — which positions the mic close to your mouth and away from noise sources — dramatically improves accuracy and costs as little as $20 to $30.
For serious daily use, a USB condenser microphone or a high-quality Bluetooth headset delivers even better results. The investment pays back quickly when you consider the time saved by fewer transcription errors that need correcting.
Software: System-Level vs. App-Specific
Speak-to-type solutions fall into two categories: those that work within a specific application and those that work system-wide. App-specific tools — like the voice typing feature in Google Docs — are limited to a single context. System-level tools work anywhere your cursor is on the Mac, making them far more practical for users who write across multiple applications.
Steno is a system-level push-to-talk app. After installation, you configure a hotkey (typically the right Option key or another convenient key), and from that point on, holding that key activates the microphone from anywhere on your Mac. Release the key and your transcription appears. There is no application to open, no mode to enter, no interface to navigate. The interaction is as close to invisible as a software tool can be.
Technique: How to Speak for Best Results
Speak in Complete Thoughts
The most important technique for speak-to-type accuracy is to compose your thought fully before you begin speaking, then deliver it in one continuous utterance. Speech recognition systems use the context of entire phrases and sentences to disambiguate words. When you speak in isolated fragments — "the meeting... um... starts at... three" — you deprive the engine of context and produce worse results than if you had spoken fluently: "the meeting starts at three o'clock."
Maintain a Consistent Pace
Speak at your natural conversational pace, not slower and not faster. A common beginner mistake is to slow down dramatically in an attempt to improve accuracy. This actually hurts results because overly deliberate speech sounds different from the natural speech patterns the model was trained on. Trust the technology at your normal pace.
Handle Punctuation Naturally
Modern speak-to-type tools can infer punctuation from the rhythm and intonation of natural speech. A rising inflection at the end of a sentence may become a question mark. A definitive downward pitch signals a period. You can also speak punctuation explicitly by saying "comma," "period," or "new paragraph," but many users find that avoiding explicit punctuation commands and letting the system handle it produces a more natural dictation flow.
What to Dictate First
Not all writing tasks are equally suited to dictation when you are starting out. These are the best use cases to build your initial speak-to-type habit:
- Slack and email replies: Short, conversational content that benefits from a natural tone and does not require precise formatting.
- Meeting notes and action items: Capturing information quickly right after a meeting while it is fresh.
- Brain dumps and outlines: Getting ideas out of your head quickly before they disappear.
- First drafts of blog posts or reports: Drafting quickly, with the understanding that you will edit later.
Avoid starting with highly technical content requiring precise formatting, code, or complex punctuation. Build fluency first on natural language content, then expand to more demanding use cases.
The Habit Formation Timeline
Most people who commit to speak-to-type report a consistent experience: the first two to three days feel slow and awkward. The first week involves frequent corrections and moments of frustration. By the second week, the cognitive overhead drops noticeably and sessions begin feeling natural. By the end of the first month, reaching for the keyboard for anything longer than a few words starts to feel unnecessarily effortful.
The key is not to give up during the first week. The awkward phase is temporary and does not reflect the long-term experience. Commit to using speak-to-type for at least 30 minutes of writing per day for two weeks, and the habit will be established.
Getting Started
You can download Steno for free at stenofast.com. Installation takes about 60 seconds and requires granting microphone and accessibility permissions so the app can insert text anywhere on your Mac. Once configured, speaking to type is as simple as holding your hotkey and talking.
The technology is ready. The only remaining step is deciding to speak instead of type — and then doing it consistently until the new behavior becomes automatic.
You have been speaking fluently since you were three years old. Speak-to-type does not teach you a new skill. It simply redirects one you already mastered into your work.