There is a persistent belief among knowledge workers that they type "fast enough." And for many tasks, that might be true. But when you look at the actual numbers — words per minute across different input methods, error rates, cognitive load, and total throughput — a clear picture emerges. Voice typing is not just marginally faster than keyboard typing. For most prose-heavy tasks, it is dramatically faster.

We have been tracking input speeds across thousands of Steno sessions, and we combined that data with published research on typing speeds, dictation rates, and error correction times. Here is what the numbers actually show.

The Baseline: How Fast Do People Really Type?

Most people overestimate their typing speed. According to a 2019 study published by Aalto University that measured over 168,000 volunteers, the average typing speed on a standard keyboard is 41.4 words per minute. Professional typists — people who type for a living — average around 65-75 WPM. Only the top 1% of typists consistently exceed 100 WPM.

Now consider speaking. The average English speaker produces between 130 and 160 words per minute in natural conversation. Even when speaking deliberately and clearly for dictation purposes, most people sustain 110-140 WPM without effort. That is a 3-4x speed advantage over the average typist, and it requires no special training.

If you want to see where you fall on the typing speed spectrum, you can try our typing speed test and compare your results against these benchmarks. Many people are surprised to find they are closer to 35 WPM than the 60 WPM they assumed.

Raw Speed Is Only Part of the Story

Raw WPM comparisons tell an important but incomplete story. What matters for productivity is net throughput — the amount of usable text you produce per unit of time, after accounting for errors, corrections, and pauses.

Keyboard typists make errors roughly every 10-15 words. Each error requires recognition (noticing the red squiggle or re-reading), cursor navigation (backspace or mouse click), and re-typing. Studies estimate that error correction adds 20-30% overhead to raw typing time. So a 40 WPM typist might only produce 30-32 WPM of clean, final text.

Voice typing with modern speech recognition engines like Whisper has a different error profile. Word error rates for clear English speech have dropped below 4% as of 2026, which means roughly one error every 25 words. But voice typing errors tend to be substitution errors — a wrong word rather than a misspelling — so they can be slightly harder to spot. We wrote more about this in our deep dive on going from 40 WPM to 150 WPM.

When you factor in error correction for both methods, the adjusted numbers look like this:

Even the most generous reading of this data shows that an average voice typist outperforms a skilled keyboard typist by nearly 2x in net throughput.

Where Voice Typing Wins Decisively

Not all text input is created equal. Voice typing shows its greatest advantage in tasks that involve producing natural language prose — the kind of work where your brain thinks in sentences and paragraphs rather than in symbols and structures.

Email and Messaging

The average knowledge worker sends 40 emails per day. If each email averages 75 words, that is 3,000 words of email text daily. At keyboard speeds, that takes roughly 97 minutes. At voice typing speeds, the same output takes about 29 minutes. That is over an hour saved on email alone, every single day.

Long-Form Writing

Blog posts, reports, proposals, documentation — any task where you need to produce 500+ words of coherent prose. Voice typing not only produces text faster, it often produces more natural-sounding text because you are literally speaking the way a reader would hear the words in their head.

Note-Taking and Brain Dumps

When you need to capture thoughts quickly — after a meeting, during research, or while brainstorming — speed is everything. Ideas evaporate fast. Voice typing lets you capture them at the speed of thought rather than the speed of your fingers.

Where Keyboard Typing Still Wins

Voice typing is not universally superior. There are legitimate scenarios where a keyboard remains the better tool.

Code and Syntax-Heavy Text

Writing actual code — with brackets, semicolons, indentation, and variable names like getUserById — is still faster on a keyboard. Voice typing works well for code comments, documentation, and commit messages, but the code itself is better typed. Steno users often develop a hybrid workflow: code on the keyboard, everything around the code by voice.

Structured Data Entry

Filling in forms, entering numbers, working with spreadsheets — these tasks involve short, structured inputs where the overhead of activating voice input outweighs any speed benefit.

Quiet or Shared Environments

This is not a speed limitation but a practical one. If you are in an open office, a library, or a shared space, speaking out loud is not always appropriate. That said, many Steno users report that speaking at a normal conversational volume — not shouting — works perfectly and does not disturb colleagues any more than a phone call would.

The Accuracy Question

People often cite accuracy as the reason they do not use voice typing. "It makes too many mistakes," they say. But this concern is largely based on outdated experience with older speech recognition systems.

Modern speech-to-text engines have reached a level of accuracy that was unthinkable even five years ago. Steno uses Groq's Whisper API, which achieves word error rates under 4% for clear English speech. For context, human transcriptionists — professional ones — typically achieve error rates between 4-6%. We are at the point where machines match or exceed human transcription accuracy.

The key factors that affect accuracy today are:

The Cognitive Load Difference

One underappreciated advantage of voice typing is the reduction in cognitive load. When you type, your brain is juggling multiple tasks simultaneously: formulating thoughts, translating them into finger movements, monitoring the screen for errors, and maintaining your train of thought. This is a significant cognitive burden that slows down not just your fingers but your thinking.

When you speak, the translation from thought to output is more direct. You think in words, and you speak in words. There is no intermediate step of mapping letters to finger positions. Many Steno users report that they not only produce text faster but that the text itself is better — more natural, more flowing, more like their authentic voice.

Building a Hybrid Workflow

The most productive approach is not to choose one method exclusively but to develop an intuitive sense for when to use each. Here is the framework we recommend:

This hybrid approach lets you capture the speed advantage of voice where it matters most while keeping the precision of keyboard input where it is needed.

Try It Yourself

Numbers on a page can only be so persuasive. The real conversion moment happens when you time yourself typing a paragraph, then time yourself speaking the same paragraph with Steno. The difference is visceral — not just a percentage improvement but a fundamentally different experience of producing text.

Start with your typing speed test to establish your baseline. Then download Steno and dictate the same passage. We are confident the data will speak for itself.