Speaking vs typing speed test
Two timed rounds over the same passage: type it, then say it. You get both speeds in words per minute, your typing accuracy, and the time the difference would save you over a thousand words.
The passage — 83 words
Dictation only helps if you can trust what lands on the screen. Most people type somewhere between thirty and eighty words a minute, and almost everyone speaks faster than that without trying. The gap is not a trick of the stopwatch; it is the reason a two hundred word email takes four minutes to type and about ninety seconds to say. Time yourself honestly, use your normal pace in both rounds, and let the numbers settle the argument. The result is rarely close.
Both rounds use this exact passage, so the two speeds are measured against the same work.
Round 1 — type it
Round 2 — say it
Press start, read the passage out loud at your normal pace, then press stop on the last word. Nothing is recorded and no microphone permission is requested — this is a stopwatch, not a recogniser.
Result
Finish both rounds and the comparison appears here.
Everything here runs on your device. Nothing you paste is uploaded.
How the test works
Both rounds use one fixed passage of roughly eighty words, shown at the top of the tool. Using the same text twice is the whole point — it removes the two things that usually make speed comparisons meaningless: different content and different length.
Round 1 starts its clock on your first keystroke, not when the page loads, so reading the passage first costs you nothing. It stops when you press “I’m done”. Your speed is your own word count divided by your own elapsed time, which means racing ahead and leaving words out does not flatter the number.
Round 2 is a stopwatch you control. Press start, read the passage aloud at a normal pace, press stop on the last word. Because the word count is known in advance, spoken words per minute falls straight out of the elapsed time.
Elapsed time comes from the browser’s monotonic clock, so a system-clock change or a daylight-saving jump mid-round cannot distort the result. Nothing is recorded: round 2 asks for no microphone permission because it is a timer, not a speech recogniser.
How accuracy is scored
Typing speed without accuracy is a meaningless number, so round 1 also scores what you typed against the source. Both texts are reduced to lowercase words with punctuation removed, and the longest common subsequence between them is measured — the longest run of words that appear in both, in the same order, allowing gaps.
The practical effect is that one dropped word costs you one word, rather than throwing the rest of the comparison out of alignment the way a position-by-position check would. Typos, extra words and skipped lines are all penalised; a different but equivalent phrasing is penalised too, because this is a transcription test.
What the numbers usually look like
Typing speeds spread widely. Most office workers land somewhere between 35 and 60 words per minute on familiar text, touch typists commonly reach 70 to 90, and copying a passage you did not write is slower than writing your own sentences because you are reading and typing at the same time.
Speaking is much less variable. Read-aloud pace clusters between 130 and 170 words per minute for most people; conversational speech runs a little faster and audiobook narration a little slower. That narrow band is why dictation changes the arithmetic so reliably: your typing speed is the variable, and it is almost always the smaller number.
A caveat the results panel repeats: this is a transcription test, not a composition test. Reading a passage aloud is the easiest possible speaking task, and typing a passage you are looking at is not the same as drafting an email. Treat the multiple as the ceiling of what dictation can save, not the number you will hit on a Tuesday afternoon.
Reading the per-1,000-word projection
Words per minute is hard to feel. The results panel converts both speeds into the time each would take for a thousand words — roughly four pages, or a decent working day of email — and shows the gap between them.
At 45 words per minute, a thousand words takes about 22 minutes to type. At 150 words per minute spoken, the same thousand takes a little over six. That difference, repeated across a week of messages, notes, tickets and replies, is where the case for dictation actually lives — not in any single dramatic moment.
Common questions
Is anything recorded or uploaded?
No. There is no microphone access, no recording and no network request. Both rounds are timers running in this tab, and closing it discards everything.
Why does my typing speed look lower than on other tests?
Most typing tests report gross WPM, defined as every five characters typed counted as one word, and many of them stop the clock the moment you reach the end of the visible text. This one counts actual words and asks you to press a button when you are finished, which is a slightly harsher measure. It is also the measure that makes the comparison with speaking valid, since speaking speed is counted in real words too.
Can I redo just one round?
Yes. Round 1 has a “Redo” button once you have finished it, and round 2 offers “Try again” after you stop the clock. “Reset both rounds” clears everything and starts over.
My speaking speed came out slower than my typing.
That happens when the reading is careful rather than natural — enunciating for a microphone that is not there, or restarting on a stumble. Try it once more at conversational pace. If it is still slower, you are an unusually quick typist and dictation’s value for you is about posture, RSI and being able to work away from the keyboard rather than raw speed.
Does dictating really run at reading-aloud speed?
Not quite. Dictating your own words includes thinking time, and a first pass usually needs a proofread. The realistic figure sits between your two measurements, and closer to the spoken one for material you already know — replies, notes, status updates, bug reports.
Where the saved time actually comes from
A speed test only measures the keyboard. In practice, dictation tools lose most of their advantage somewhere else: in a separate window you have to copy out of, in a transcript you have to punctuate by hand, in the “um”s you have to delete afterwards. Software that is 3× faster to speak into and then costs two minutes of cleanup has given the time straight back.
VoiceSnap Pro is built around that arithmetic. You hold one keyboard shortcut, speak, and clean punctuated text appears in whatever field your cursor is already in — your email client, Slack, a code editor, a browser form. Punctuation and paragraph breaks are automatic, filler words are stripped as you go, and every dictation is saved to a searchable notes library in case you need the original back.
It runs on macOS and Windows, handles 50+ languages including switching mid-sentence, supports voice commands for formatting, and accepts a custom vocabulary for the names, jargon and acronyms that generic models mangle. Nothing you dictate is used to train AI models, and it is a one-time purchase rather than a subscription. The app has not shipped yet — join the waitlist to hear on release day.
Want to keep measuring? The speech time calculator turns a script into minutes at your own pace, the word counter gives you length and speaking time as you write, and voice typing lets you try browser dictation without installing anything.