Capgemini Exceller: the test pattern, and what to prepare for
Capgemini calls Exceller its flagship program for technical students, run through engineering institutes and graduate colleges across India, and its pages send candidates to their college placement team rather than describing a test. What the assessment contains is therefore entirely a matter of report, and the reports are unusually consistent: an online assessment with a technical section of multiple-choice questions and pseudocode, an English communication test, a game-based cognitive round, and a behavioural assessment, followed by a coding round and interviews.
The game-based round is what makes Exceller different from the other mass-recruiter papers on this site. It replaced a conventional reasoning section a few years ago, and it is not something a question bank prepares for. This page is honest about that: it maps the technical and coding parts to the December Code subjects that cover them, names the games, English and behavioural parts as outside the site's scope, and ends with a practice set from pages whose samples are already public.
December Code is not affiliated with, endorsed by or connected to Capgemini. This page summarises how Capgemini Exceller is publicly described, from the sources listed at the end, checked on 29 September 2026. Where a detail comes from prep sites or candidate reports rather than Capgemini itself, the page says so. The notification or test invitation for the drive you are sitting is the source of truth, and where it differs from this page, it is right.
How the process runs
A round is marked official when Capgemini describes it on its own pages, and commonly reported when the description comes from prep sites and candidate reports.
- 1
Application through the college placement teamofficial
Capgemini's technical-graduates page says candidates join the hiring process through their college placement team. Off-campus Exceller drives are announced on Capgemini's India careers pages when they run.
- 2
Exceller online assessmentcommonly reported
Commonly reported as four parts taken in one sitting under AI proctoring: technical MCQs and pseudocode, an English communication test, a game-based cognitive test, and a behavioural and power-skills assessment. Reports say each part has its own threshold and that failing any one ends the process regardless of the others.
- 3
Coding roundcommonly reported
Reported as two problems in about 45 minutes on arrays, strings, linked lists, stacks, queues, trees and simple dynamic programming, with a bar expressed as a share of test cases passed. Some reports place it inside the first sitting, others as a separate round.
- 4
Technical and HR interviewscommonly reported
Candidates report a technical interview on the project, a programming language, object-oriented concepts, SQL and operating-system basics, then an HR conversation about relocation and the training period. Capgemini's page describes the training that follows: an eight-to-ten-week central program.
What the test asks, and what prepares you for it
2 of 5 sections map to a December Code subject. The rest are named below, not glossed over.
| Section | What it tests | Prepare with |
|---|---|---|
| Technical MCQs and pseudocodecommonly reported | Programming logic in C, C++ and Java, data structures, object-oriented concepts, DBMS and SQL, operating systems and networks, and pseudocode tracing. Commonly reported as 40 questions in 40 minutes with a threshold around 60%. | OOPDBMSSQLOperating SystemsComputer NetworksDSACJava |
| English Communicationcommonly reported | Grammar, sentence correction, reading comprehension and vocabulary. Commonly reported as 30 questions in 30 minutes with a threshold around 65%. | |
| Game-based Cognitive Testcommonly reported | Four short interactive games drawn from a larger pool, reported to measure pattern recognition, working memory, deductive logic and processing speed. Around 20 to 30 minutes. | |
| Behavioural and Power Skillscommonly reported | Situational and preference questions on teamwork, decision-making, adaptability and stress. About 20 minutes; included in most drives, not all. | |
| Codingcommonly reported | Two problems in about 45 minutes, in the mainstream languages, graded on test cases: arrays and strings most often, linked lists, stacks and queues, and the occasional tree or dynamic-programming problem. | DSAJavaCPython |
What the games are, and what they are not
Reports describe the cognitive round as four games chosen from a pool of a couple of dozen, with names such as a grid challenge, a digit challenge and a motion challenge, each a few minutes long. They test how fast you notice a pattern, hold a sequence in memory and apply a rule, not how well you know percentages. Two things follow. First, the round cannot be prepared for with a question bank, this one included; the practice that helps is the games' own practice modes and a calm first attempt. Second, reports say the games carry their own threshold and are never optional, so a strong technical score does not rescue a weak game score.
The conventional aptitude that December Code measures is therefore less central to Exceller than to the TCS, Wipro or Cognizant papers. Where it does apply is the technical section's pseudocode, which is logic, and the interview's reasoning questions.
Where the CS core comes in
The technical section is the part of Exceller that a computer-science student can prepare for directly, and reports describe it as broad: a short program to trace, a question on which sorting algorithm is stable, a primary-key or normalisation question, a deadlock or paging definition, a networking layer, an object-oriented concept. The skill is breadth under a minute a question. Every one of those areas has a hub on this site, and a diagnostic in each shows which would cost you marks.
The coding round is reported at the easy-to-medium level, with problems that need a loop and a condition more often than an algorithm. Reports mention linked lists, stacks and queues among the problem types, which puts it a step above the array-only coding rounds of some other drives, and describe a pass mark stated as a fraction of test cases rather than all of them.
What moves between drives
One prep site that describes the assessment in detail also says, in as many words, that Capgemini's current India recruitment page gives no fixed Exceller paper, that the pattern varies by campus and role, and that the frameworks being discussed for the 2027 batch, with AI literacy and debugging, are not the pattern the 2026 batch sat. That matches what Capgemini's own pages show: a program description and a pointer to the placement team, not a syllabus. So the invitation your placement cell forwards is the specification; this page is the shape to expect and the subjects to prepare.
What December Code covers here, and what it does not
The diagnostic and practice cover DSA, OOP, DBMS, SQL, Operating Systems, Computer Networks, C, Java, Python: the subjects behind 2 of the 5 sections above. It does not cover the following, and no page here pretends to:
- The game-based cognitive round: interactive games measuring pattern recognition, memory and speed. No question bank prepares for it, and December Code does not claim to.
- The English communication test and the behavioural assessment: two more parts of the online assessment outside the site's scope.
- The coding judge: two problems executed against test cases. December Code prepares the data-structure reasoning in MCQ form and does not run your code.
What could not be confirmed from Capgemini
- Every count, duration and threshold on this page (40 technical questions, 30 English, four games, 20 minutes behavioural, two coding problems in 45 minutes, thresholds of 60% and 65%) is a prep-site report; Capgemini publishes none of it.
- Whether the behavioural assessment and the coding round are part of every drive. Reports differ.
- How the game-based round is scored.
A practice set for Capgemini Exceller
Topic pages in the subjects this process tests, each with the questions interviewers ask and one question from the bank exactly as the diagnostic asks it. Nothing is published for this page: every sample was already public on its topic page.
- Sorting algorithm interview questionsSorting algorithms and their properties are reported technical MCQs.
- Linked list interview questionsLinked lists are named among the coding round's problem types.
- Stack and queue interview questionsStacks and queues are the other named coding structures.
- Array and two-pointer interview questionsMost coding problems are still arrays with a condition to satisfy.
- String interview questionsString problems are the second most reported coding family.
How the diagnostic asks it
One public question per subject, the same one its hub shows. The answer is marked because these are public; in a sitting you choose first, then see why each option is right or wrong.
Which of the following operations on an unsorted array of n integers takes O(n) time in the worst case?
- 1Finding the maximum elementcorrect
- 2Appending an element at the end when spare capacity already exists
- 3Reading the element at index k
- 4Reading the array's length
In an unsorted array the maximum could be anywhere, so every element must be examined: O(n). Indexed reads are O(1) because the address is computed from the base and the index; appending into existing spare capacity just writes one slot, O(1); and the length is stored, so reading it is O(1). Only a sorted array would let the maximum be read directly.
What is 'inheritance' in object-oriented programming?
- 1A mechanism where multiple methods share the same name but different parameter lists.
- 2A mechanism where a new class acquires the properties and behavior of an existing class.correct
- 3A mechanism where a class is prevented from being instantiated directly.
- 4A mechanism where an object's internal data is hidden from other classes.
Inheritance lets a derived (child) class reuse and extend fields/methods of a base (parent) class, promoting code reuse and establishing an 'is-a' relationship. Several methods sharing a name with different parameter lists is overloading; hiding an object's internal data from other classes is encapsulation; and preventing a class from being instantiated directly describes an abstract class.
Measure it
Knowing the pattern tells you what is coming. A diagnostic tells you what you would get wrong.
10 DSA questions across its topics, easy to hard, about fifteen minutes. You get a readiness figure with the arithmetic shown, the topics you missed named, and a practice set sized for today. Free: 1 diagnostic a month and 15 problems a day. No card.
Sources, and when they were checked
Pattern last verified 29 September 2026, against the pages below. Patterns change between drives and cycles; when a source changes, this page changes and the date moves with it. A source marked official is Capgemini’s own; the others are the established prep sites and candidate reports the “commonly reported” details come from.
- officialCapgemini India: Technical graduates (Exceller)· what Exceller is, the route through the college placement team, and the training that follows
- officialCapgemini India: Students and graduates· the graduate programs page; no assessment details
- prep sitePlacement Preparation: Capgemini syllabus and test pattern 2026· the section counts, durations and thresholds reported for the online assessment and the coding round
- prep siteGUVI: Capgemini Exceller assessment 2026· a second report, with its own caveat that Capgemini gives no fixed paper and that 2027-batch frameworks differ; dated 24 September 2026