Hindi Typing, InScript, CPCT, Government Jobs, India
Most people are taught Hindi typing the wrong way. Here is how WPM, KDPH and accuracy are actually calculated, and what the InScript layout does to your fingers.
## I built a free Hindi typing test because every other one wanted my login
Last month I needed to check one number. My Hindi typing speed. I typed for a passage, got a score, and wanted to know whether it was any good.
Then the tool asked me to sign up.
I closed the tab. That was the whole moment. Somewhere between a browser tab and a scoring sheet, a free utility had quietly become a lead generation form, and I was not willing to trade an email address for four keystrokes per second.
So I built one that asks for nothing. It opens, you type, you get your numbers. That is the entire product. It is live at zettool.com if you want to skip ahead and try it.
This post is about what I learned building a Hindi typing speed test, because the interesting part was not the test itself. It was how much of Hindi typing advice on the internet is quietly built on wrong assumptions.
If you searched here for how to improve Hindi typing speed, or what a good Hindi typing speed in Mangal font actually is, the sections below answer both.
## The number everyone measures wrong
Here is the first thing that surprised me. Government typing tests do not count words. They count keystrokes, and they call it words.
Five keystrokes equals one word. Long word or short word, the counter does not care. An eight letter word and a two letter word each move the score by one point if you typed five keys’ worth of them.
Once you internalise that, the scoring stops being mysterious:
**Gross WPM** = (characters ÷ 5) ÷ minutes
**Net WPM** = Gross − (mistakes ÷ 5) ÷ minutes
**Accuracy** = correct characters ÷ total characters typed
That middle line is the one that changes behaviour. Every single mistake costs five keystrokes, which is a whole word off your score. It is not a small deduction that smooths your score down over time. It is a cliff.
This is why someone typing 40 WPM at 75 percent accuracy will fail a test that someone typing 35 WPM at 94 percent passes. Both finish at a similar net speed, roughly 30 to 33 WPM, but only one of them clears the accuracy requirement. The faster number looks better on paper and gets rejected in the room.
## SSC uses a different number entirely
If your exam is SSC CGL DEST, none of the above matters, because that paper is not scored in WPM at all. It is scored in KDPH, key depressions per hour, where the usual qualifying figure is 8000.
8000 KDPH works out to roughly 27 WPM.
So two candidates can both be told “you need 35 WPM” and one of them has been given the wrong unit for their exam. The result card I built shows WPM, KDPH and CPM side by side, so at least the person taking the test can see which scale their score is actually being read on.
For a CPCT typing test the figure usually quoted is 30 net WPM in Hindi, but read your own latest notification, because the requirement changes between cycles.
## The InScript layout is genuinely hostile
Get Nayab Mirza’s stories in your inbox
Join Medium for free to get updates from this writer.
Here is the part that made me respect the people who already type Hindi well.
The Windows InScript keyboard layout puts the vowels across the top row in alphabetical order, which sounds helpful and is not:
- **D** produces अ
- **Shift + D** produces ् — the halant, or virama
So the single most common letter in the language sits under a key whose shifted form is a combining mark that joins letters together. Every consonant word you type ends in a halant that you never think about consciously but that your right pinky has to find every single time.
It gets stranger further right. **M** is स, period is the Devanagari danda ।, and slash is य. The punctuation row is not where English muscle memory says it is.
This is also why Mangal font and Kruti Dev are not interchangeable. Kruti Dev uses a Remington layout where the keys are positioned completely differently, so someone who practised for months on the wrong one has to start again. Check which font your exam notification names before you type a single paragraph.
I printed the whole layout out and taped it above my monitor to build the tool’s keyboard guide. That guide is on the site too, at zettool.com/keyboard, and it is the part of this project I would defend hardest. Anyone learning Hindi typing is really learning an unfamiliar geography, and no amount of general advice about touch typing prepares you for that.
## Backspace is a different skill, not a smaller one
Practice mode lets you fix mistakes. The exam does not.
That sounds like a small difference. It is not. In practice mode you learn to type correctly. In exam mode you learn to keep moving, which is a separate skill, and the only way to get it is to practise without the crutch.
This is where Hindi typing paragraph practice matters more than it looks. Repeating one passage you have already memorised makes your speed look good and teaches you nothing, because the exam gives you text you have never seen. That is why the test hands you a different paragraph on every reset.
So the test has two modes for a reason. Practice mode is where accuracy gets built. Exam mode locks backspace, runs the full timer, and shows the result only when time is up, because that is what the test day feels like.
If you only ever practise in practice mode, you are training a different muscle than the one being examined.
## What I did not build, on purpose
No accounts. No certificate generator. No leaderboard. No progress tracking in the cloud.
Each of those is a feature I could have added in a weekend, and each one would have required me to know who you are. The whole point was that this stays a page you open and close. It runs entirely in your browser, your typing never reaches a server, and there is nothing to delete afterwards.
I mention this because free tools on the internet are almost never really free. The cost moves somewhere. Usually it becomes a signup, an ad, or a dataset you did not know you were giving away.
## The part I still do not know
I do not know if it helps anyone yet.
I am not going to pretend this post has a lesson about growth or audiences. It is a typing test that does exactly one thing, and the honest position is that I have built it, shipped it, and I do not yet have evidence that anyone is using it.
What I can say is that it took a weekend, and the people it helps are the ones who needed a number and did not want an email address for it.
If that is you, the test is at zettool.com. Run a one minute Hindi typing test, then run a ten minute exam mode one, and write the accuracy down. Tomorrow, do it again.
Hindi Typing, InScript, CPCT, Government Jobs, India
