What the TCS NQT actually tests, and why the score band matters more than passing

On 6 September 2026 I tried to open TCS’s own NQT page four different ways: the careers portal, the early-careers page, tcsnqt.com, and the press release announcing the test. Three came back HTTP 403. The fourth loaded blank. That’s not a reflection on TCS’s exam, just a note that if you’re hunting for the official syllabus PDF, expect to log into the candidate portal rather than Google your way there.

So this guide leans on two exam-prep sites I fetched the same day, testbook.com and hirist.tech, and flags every spot where their numbers disagree. TCS NQT decides more than whether you get hired. It decides which of three very differently paid jobs you get, and most candidates don’t find that out until after the results are already in.

What TCS NQT is, in one paragraph

TCS NQT, short for TCS National Qualifier Test, is the entry gate Tata Consultancy Services runs for campus and off-campus freshers across three outcomes: Ninja, Digital, and Prime. It’s a single test. Where you land afterward depends on your percentile, not on which role you ticked at registration. You don’t pick Prime. Prime picks you, if the score supports it.

The 2026 cycle opened for graduates from the 2024, 2025, and 2026 passing years, with the application window closing around 2 September 2026 for that round, according to testbook.com’s listing checked the same day. TCS runs multiple cycles a year, so if you’ve missed one, there’s usually another within a few months.

The sections, and where the two sources disagree

Here’s where I have to be honest about the gap. Testbook.com describes a Foundation and Advanced structure totalling 180 minutes and 92 questions, split into cognitive sections (verbal, reasoning, numerical) plus a programming logic and hands-on coding block. Hirist.tech, fetched the same afternoon, describes the same two-part shape but puts the total at 190 minutes and roughly 90 questions, with a heavier advanced coding block of 90 minutes for 3 questions. Neither number is wrong exactly; they’re probably describing slightly different cycles or rounding the same schedule two different ways, and TCS doesn’t publish a single canonical version anywhere I could load.

Ten minutes and two questions isn’t a huge gap, but it’s exactly the kind of detail that changes if TCS tweaks the pattern between cycles, and neither source is TCS’s own copy. Treat the following as directionally right, not gospel:

There’s no negative marking, both sources agree on that, which changes your strategy: guess rather than skip. A blank answer and a wrong answer cost the same, so leaving 15 questions unattempted because you ran out of confidence is just leaving points on the table.

Time pressure hits the coding block hardest. Ninety minutes for two or three problems sounds generous until you spend the first twelve minutes just parsing the problem statement, which happens more often than candidates admit afterward. Read the constraints section first, every time, before you write a line. It’s a small habit and it saves real minutes.

Ninja, Digital, Prime: what the gap actually looks like

This is the part nobody explains clearly enough before the exam. Hirist.tech’s breakdown puts Ninja around the 50th to 60th percentile with UG pay near 3.3 to 3.6 LPA, Digital around the 65th to 75th percentile at roughly 7.09 to 7.72 LPA, and Prime at the 80th percentile and above, paying about 9.09 to 9.66 LPA. That’s close to triple the Ninja number for Prime, from the same exam, on the same day, for candidates who might have identical resumes.

I’d treat those exact percentile cutoffs as approximate and expect them to shift by cycle and by how many people take the test that season. What doesn’t shift is the mechanism: a marginal 8 or 9 extra correct answers can move you from Digital pay to Prime pay. That’s not a reason to panic. It’s a reason to spend your last week on accuracy in the sections you’re already decent at, rather than trying to fix your weakest one from scratch.

Worth saying plainly: none of this means Ninja is a bad outcome. It’s still a job offer with a defined salary and a real career ladder inside TCS, and plenty of people move up internally within a couple of years regardless of which door they walked in through. The percentile gap matters for your first paycheck, not necessarily for your fifth year.

Eligibility, and the one question neither source answers

Eligibility, per both sources, sits around 60% or a 6 CGPA across academics, no active backlogs, age 18 to 28, and up to a 24-month academic gap tolerated. Freshers with up to two years of prior work experience can also sit the test, which surprises people who assume NQT is campus-only.

One thing worth flagging: neither source states whether the percentile is calculated against everyone who registers or only against everyone who actually shows up and finishes the test. Those are different denominators, and the gap between them could plausibly matter more than a point or two on your raw score. I asked around and got three different answers from three different prep forums, which tells you the candidate side doesn’t actually know either.

A 5-week prep plan that isn’t the same generic advice everywhere

Most prep guides tell you to “practice aptitude daily” and leave it there. Here’s a version with actual allocation, built around the fact that coding carries more weight per minute than the cognitive sections do.

Week 1: Diagnostic. Take one full mock under real time pressure, no notes. Score every section separately. You need to know if your bottleneck is verbal (fixable in days) or coding logic (needs weeks).

Week 2 and 3: Split time roughly 60/40 toward whichever section scored worst, but never drop a section to zero. A verbal score of 40% pulls your overall percentile down even if coding is strong, since the sections aren’t weighted equally in your favor just because one feels more important.

Week 4: Coding-specific. Two problems a day, timed at 15 minutes each, not the 45 minutes you’d give yourself solving them for fun. NQT coding rewards speed on medium problems over brilliance on hard ones.

Week 5: Two more full mocks, three days apart, then rest for the last two days. I don’t have data on whether resting actually improves scores versus one more day of grinding. I suspect it does, mostly because panic-cramming the night before tends to wreck sleep, and a tired brain reads slower on a 25-minute verbal section than a rested one does.

If you’re prepping the softer side of the process too, our HR round questions for freshers post covers what comes after NQT clears, and notice period at TCS, Infosys, Wipro and Accenture is worth reading before you accept, since services-company notice periods surprise a lot of first-time joiners.

What isn’t stated officially, and what I’d stop worrying about

Score cutoffs, retake limits per cycle, and whether TCS NQT scores carry over to partner companies beyond TCS: none of that showed up in an official TCS document I could actually load on 6 September 2026. Hirist.tech claims there’s no fixed attempt limit within the age window, which sounds plausible but isn’t something I can confirm against TCS’s own text, so treat it as a rumor with decent provenance rather than a rule.

A related question people ask us often: does a low TCS NQT score this cycle hurt a reattempt next cycle. Neither aggregator addresses this directly, and I’d rather say that plainly than guess at an answer that sounds confident but isn’t grounded in anything I actually checked.

What I’d stop worrying about: the exact minute split between sections. Whether verbal gets 25 or 30 minutes changes nothing about how you should prepare. Spend that anxiety on the coding block instead, since that’s where the Prime-versus-Digital gap actually gets decided, not in the reading comprehension paragraph about rainfall patterns.

Admit cards, when TCS finally sends them, usually list your exact slot and section order a few days ahead. That’s the point to stop guessing and start trusting the document in front of you over anything an aggregator, including this one, told you weeks earlier.

Once you’re through TCS NQT and into interview rounds, mock practice against a real timer catches more mistakes than another round of aptitude questions does. Coding interviewers notice hesitation faster than they notice a wrong answer, oddly enough, and a rehearsed explanation reads better than a correct-but-silent one. That’s the whole premise behind our interview email templates tool and the practice sessions inside LastRound AI, if you want a second pair of eyes before the actual call.