Is speaking Chinese faster than typing pinyin?

In the one controlled study we could open, speaking was faster for short messages. Mandarin entered by speech reached 123 words per minute, against 43 words per minute on the built-in pinyin keyboard of an iPhone ^{1 2}. That is 2.87 times faster. The result is a good reason to try voice input, but it is not a measurement of your own typing, and it does not tell you how long you will spend correcting a dictated draft afterward.

The phone study shows a gap of almost three to one

The study had participants transcribe short messages twice, once with speech recognition and once with the phone's pinyin keyboard. Speech reached 123 words per minute, and the keyboard reached 43 ¹. The English condition of the same study counted the finished text, including time spent correcting errors ³, and the Mandarin condition used the same laboratory protocol ¹.

The study ran the same design in English. Speech reached 153 words per minute and the keyboard reached 52, a ratio of 2.93 ^{3 4}. The Mandarin and English gaps are close, so the advantage of speech is not special to Chinese.

The speed numbers apply to the study's conditions. Its typists used an iPhone, and the speech used an older recognizer, Baidu's Deep Speech 2 ¹. A newer recognizer may be faster or more accurate, and a faster keyboard may narrow the gap.

Why the Mandarin gap may not match your own use

Three features of the study limit how far its numbers reach. First, the messages were short and the setting was a laboratory ¹. Second, pinyin speed depends on the typist: the keyboard result depends heavily on experience with choosing candidates from the suggestion list ¹. Third, the keyboard figure reflects its word suggestions, so the number says little about other Chinese input methods ².

Chinese adds a fourth issue, which is what a "word" is. Chinese writing has no spaces between words, so a words-per-minute figure depends on how the text is split. Native speakers agreed on where word boundaries fall only about 75 percent of the time in one early study ⁵. Compare speed figures only when they come from the same counting method.

Where the answer depends on your situation

• Desktop keyboards. The study measured typing on a phone. We did not find a published desktop study of pinyin speed in the sources we opened, so a desktop typist may do better than the 43 figure suggests.
• Quiet rooms. The English version of the study was run in a quiet room with short, read phrases ³. Speech in a noisy office or on a street may be slower or less accurate, so test in the place you will use it.
• Editing. Dictating a draft and then correcting it takes different time from dictating a clean message. The study does not measure long drafts or the editing that follows them.
• Precision. When you need exact characters, choosing each one from a pinyin list can be more reliable than checking a dictated sentence, so a keyboard may be the better tool for some text.

What to do

1. Time yourself on your own text. Pick a paragraph you normally write, then dictate it and time the attempt. Type the same paragraph on your usual keyboard, and time that too. Include the time you spend fixing mistakes, as the English condition of the study did.
2. Use voice for first drafts. Speech may be quicker when you are composing, because you do not stop to choose characters.
3. Use the keyboard for fine edits. Corrections to single characters may be quicker to type.
4. Check the output for homophones. Mandarin has many syllables that map to several characters, so read each dictated sentence before you send it. For the term, see pinyin.

For how to set up Chinese voice input on each system, see how to voice type in Chinese. For the wider comparison of speech and typing in Asian languages, see the guide to Japanese, Chinese and Korean speech to text in the Languages section.

Related answers

• How do you voice type in Chinese?
• How do you voice type in Korean?
• What is the best speech to text for Japanese?
• How fast do people type in Japanese?