Hindi typing test

Type Hindi you would actually write — the 600 most frequent words in the language, not a passage about the quick brown fox.

15 left
No keyboard installed? Click the keys.

The three ways people type Hindi, and why your speed differs on each

Almost every Hindi typing question online is really a layout question in disguise. There are three systems in common use and they do not share muscle memory, so a score from one tells you very little about your speed on another.

InScript is the layout the Government of India standardised, and it is what Windows installs when you add Hindi. Vowels sit on the left hand, consonants on the right, and the arrangement follows the traditional Devanagari alphabetical order rather than letter frequency. Its real advantage is that the same key positions work across Indian scripts — the Devanagari InScript mapping is identical for Hindi and Marathi, so a Marathi typist can sit down at a Hindi machine and type immediately. InScript produces Unicode, which is what this test measures.

Remington GAIL reproduces the key arrangement of the mechanical Hindi typewriter. It exists because a generation of typists learned that layout before computers, and offices kept it. It is typically paired with Kruti Dev (often Kruti Dev 010), and this is where people get caught out: Kruti Dev is not a Hindi encoding at all. It is a font trick. The file underneath holds ordinary Latin characters, and the font simply draws Devanagari shapes over them. Change the font and your document turns back into gibberish. The text cannot be searched, cannot be pasted into a form, and is invisible to anything expecting real Hindi.

Phonetic input — typing namaste and getting नमस्ते — is how most people type Hindi on a phone. It is the fastest system to start with and the slowest to be precise in, because the transliteration engine, not you, decides the spelling.

This test uses Unicode Devanagari. If you practise here and then sit an exam that requires Remington GAIL with Kruti Dev, your hands will be in the wrong places — that is a layout difference, not a change in your ability.

What this word list actually measures

Hindi is the shortest-word language of the three Devanagari tests here, and still nowhere near a Latin script. Everything below is counted from the exact list this test serves.

600words in the test list
2.55visual units per word
4.36codepoints per word
310distinct letter shapes
90%words that stack

The gap between those middle two numbers is the whole reason this test exists. A word here takes about 4.36 keystrokes to build but occupies 2.55 visual units on the page — a ratio of 1.71 to one. On this word list that convention would report roughly 1.7 times the words-per-minute figure shown here, because it counts the keystrokes hidden inside conjuncts as if they were separate letters.

Practising Devanagari, as opposed to practising typing

The usual advice — do a test every day, watch the number rise — misses where Hindi typing actually breaks. Run this test a few times and look at the missed-character list on the result card rather than the speed. It will almost certainly be matras and half-letters, not consonants. Consonants are one key each and they were never the problem.

So drill the thing that is failing. Pick the twenty words in your last run that contained a conjunct, and type those slowly until the sequence stops needing thought. क्ष, त्र and ज्ञ appear often enough in ordinary Hindi that fluency in those three alone moves your whole score, and they are three keystrokes each rather than one.

Then work on duration rather than speed. Fifteen seconds measures your peak and nothing else; the useful reading is whether two minutes looks like your fifteen seconds. If it drops sharply, the limit is concentration and hand tension, not layout knowledge, and more speed drills will not fix it.

About the exam numbers you have seen elsewhere

Search for a Hindi typing requirement and you will be told 20 words per minute, or 25, or 30, or 40 — frequently for the same exam, on the same day. The numbers are not all wrong. They are answers to different questions.

Madhya Pradesh's CPCT is the one most people mean, and it does not work the way a pass mark works. Typing is scored on a scale rather than marked pass or fail, the result goes on a score card valid for seven years, and each recruiting department then decides what it wants to see on that card. That is most of the explanation for the contradictory numbers: a site quoting "30 WPM Hindi" is quoting one vacancy notification, and a site quoting 20 is quoting another.

There is a floor underneath all of it, though. Secondary sources agree that Hindi below 20 WPM — or English below 30 — scales to zero rather than to a low score, which makes 20 a threshold and the higher figures department cut-offs sitting above it. That could not be confirmed at CPCT's own portal, which does not answer requests from outside India, so treat it as the first thing to check in the notification for your post rather than as settled.

So take the number from the notification for the post you are applying to, and treat every figure on a typing-practice site — this one included — as something to verify rather than rely on. What a practice test can honestly give you is a consistent measurement over time, which is the thing that actually tells you whether the practice is working.

Where Devanagari typing actually goes wrong

ि

The vowel sign that moves

In कि you see ि before the क, but you type क first and the ि second. The script reorders it for you. Typists coming from the Preeti/Kruti Dev world type it the other way round out of habit and get कि wrong every time.

्

The virama that builds conjuncts

क्ष is three keystrokes — क, virama, ष — and lands as one visual unit. That gap between what your fingers do and what the screen shows is why counting keystrokes and counting letters give such different speeds in this script.

़

The nukta letters

क़ ख़ ग़ ज़ ड़ ढ़ फ़ each exist twice in Unicode: as a single codepoint, and as the base letter plus a nukta. They look identical and compare as different strings, which is how a correct-looking word gets marked wrong.

ं

Anusvara against the half-letter

हिंदी and हिन्दी are both current spellings. This test accepts what its word list holds, which is whichever form the frequency corpus recorded.

Sources

The layout claims above — InScript as the government standard, Kruti Dev as a font mapping rather than an encoding — were checked rather than recalled. The exam portals below are the primary ones; neither answers requests from outside India, so the 20 and 30 WPM floors are reported from secondary sources and flagged as such above rather than asserted.

If you came here to write, not to be measured

This page measures. It is not where you write. The Hindi keyboard has the full layout, predictive text and an editor that will not vanish when the timer stops — and it draws on the same frequency list the test above uses, so the words it suggests are the words you just practised. Other scripts are on the typing test hub.

FAQ

Frequently Asked Questions

How is the word count calculated here?
Two ways, shown side by side. The main figure counts whole words you typed correctly. The second counts visual letter units — grapheme clusters — per minute. Most typing tests use neither: they divide characters by five, an English convention. On this word list that convention would report roughly 1.7 times the words-per-minute figure shown here, because it counts the keystrokes hidden inside conjuncts as if they were separate letters.
Why is my speed lower here than on an English typing test?
Partly the script and partly the measurement. A Hindi word in this list averages 2.55 visual units against 4.36 codepoints, so the same physical effort produces fewer finished words. Hindi typists are not slower; Devanagari words carry more per unit.
Do I need to install a Hindi keyboard to take this test?
No. The on-screen Devanagari keyboard below the test is the same key grid used on our Hindi keyboard page, and you can click it. If you do have Hindi installed, type normally and ignore it.
Which layout is the on-screen keyboard?
It is arranged for direct character entry rather than reproducing InScript or Remington key-for-key, because it has to work for people who have never installed either. For InScript muscle memory, practise on a real InScript layout.
Where do the test words come from?
A frequency list built from a real Hindi corpus — the same verified list our Hindi keyboard uses for autocomplete. It was not written by hand, which is why it contains the words Hindi actually uses most rather than the words someone thought it should.
Is Kruti Dev typing measured here?
No, and it could not be. Kruti Dev text is Latin characters wearing a Devanagari font. There is nothing for a Unicode test to read. If your exam requires Kruti Dev, practise in the software your exam uses.
What is a realistic Hindi typing speed?
It depends entirely on which of the three layouts you learned and how long ago. Rather than compare yourself to a number from a stranger, take this test now, then again in two weeks. Your own trend line is the only comparison that means anything.
Does the test count my mistakes?
Yes. Accuracy is shown with the speed, and the result card lists which characters you missed most often — usually the matras rather than the consonants.
Why do some words look longer than they are?
The longest word in this list, अलग-अलग, is 7 visual units but more codepoints than that. Conjuncts and matras stack invisibly.
Can I practise for CPCT here?
You can practise Hindi typing here. CPCT is sat in specific software with a specific layout, so treat this as speed and accuracy practice rather than an exam simulator.
Does it work on a phone?
Yes, using your phone keyboard. Phone Hindi input is usually phonetic, so your speed there measures something different from a desktop layout.
Is anything saved?
Your best score is kept in your own browser and never leaves it. Clearing your browser data clears it.
Is Hindi InScript the same as Marathi InScript?
For the shared letters, yes — the Devanagari InScript mapping is the same. Marathi adds ळ and a few forms Hindi does not use, which is why the two tests here use different word lists rather than the same one relabelled.
Copied!