Capgemini Exceller Assessment 2026: Complete Framework Guide

Capgemini hires freshers across India through its Exceller program, and the 2026 assessment is now a five-stage framework, not a single aptitude test. If you are prepping from an older Capgemini guide, the shape has changed. Studying the wrong sections is the most common way candidates lose time.

This guide maps the framework end to end. It covers the five stages, what each one tests, how elimination works, the package tiers, and where to start. Treat the numbers here as the structure of the test. Exact section counts and timings shift from one campus drive to the next, which Capgemini itself notes in its recruitment process guidance.

What the Capgemini Exceller assessment is

Exceller is Capgemini’s campus hiring program for technical graduates, described on the company’s technical graduates page as the route through which it recruits emerging engineering talent from institutes across the country. You apply once, usually through your placement cell or the Superset platform, and the assessment decides both whether you are selected and which package tier you land.

The important mental model is simple. This is a tiered evaluation, not a pass-or-fail gate. Two candidates can both clear the process and still receive very different offers. The separator is mostly coding performance in the later stages.

Capgemini runs Exceller as one of the larger fresher pipelines in Indian IT. That scale is why the framework is standardised into stages, and why knowing the shape in advance is worth more than raw last-minute cramming.

The five-stage Exceller assessment journey

The 2026 framework Capgemini has shared with candidates groups the assessment into five stages. Several are eliminatory, so you clear one to reach the next.

Stage	What it covers	Format
1. English Communication	Listening, speaking, reading, writing	AI-scored language tasks
2. Technical Module	AI Literacy plus technical (DSA, SE fundamentals)	Multiple-choice questions
3. Debugging Assessment	Find and fix issues in given code	Code, C / C++ / Java
4. AI-Assisted Coding	Solve a problem by directing an AI assistant	Guided, step-by-step task
5. Cognitive Assessment	Working memory, reasoning, behaviour	Game-based and adaptive modules

The progression rule is the part candidates underestimate. A few stages are eliminatory. A brilliant communication score cannot rescue a failed technical module. Consistent readiness across all five beats a spike in any one.

What changed from older Capgemini guides

Older Capgemini test guides describe a pseudocode-plus-quant-plus-English online test followed by a coding round. The 2026 framework keeps the spirit of that structure but re-labels and expands it. Two additions stand out. The technical module now carries an explicit AI Literacy section, and a separate AI-assisted coding stage has appeared that no other India IT recruiter currently runs. If your prep notes predate this, you are missing the two newest, lowest-competition parts of the test.

Stage-by-stage: what each round tests

English Communication

The first stage checks listening, speaking, reading, and writing through an AI-scored module. Public candidate reports describe a spoken component where you speak on a prompt for a short window, judged on pronunciation, fluency, and clarity. It rewards steady, clear speech over a rehearsed accent.

Technical Module

This is the biggest change from older Capgemini guides. The technical module now pairs an AI Literacy section with a classic technical section. AI Literacy covers foundations of generative AI, prompt basics, and responsible-AI awareness. The technical section stays close to what Capgemini has always tested: programming logic, data structures and algorithms, and software-engineering fundamentals such as DBMS, SQL, and version-control basics.

Debugging Assessment

Here you read a problem and a block of code that already contains one or more issues, then correct it so the program runs as expected. The languages reported by candidates are C, C++, and Java. The skill is reading someone else’s logic and spotting where it breaks, which is different from writing fresh code.

AI-Assisted Coding

This stage is distinctive to Exceller and has almost no equivalent in other India IT drives. Instead of writing a full solution alone, you direct an AI assistant through a step-by-step flow, giving clear inputs at each step to reach working code. Capgemini’s stated focus is the process: how well you frame a problem and guide the tool, not whether you paste a final answer.

Cognitive Assessment

The final stage uses game-based modules for working memory, spatial attention, and logical reasoning, plus an adaptive behavioural module. Our Capgemini game-based aptitude guide covers this format in detail, including why timed practice beats memorising game walkthroughs.

Eligibility, packages, and the elimination rule

Eligibility for Exceller is broadly consistent across drives, and the Capgemini selection process guide breaks down the finer points. The commonly reported bar:

• B.E, B.Tech, M.E, M.Tech, or MCA, with M.Sc in CS or IT often eligible.
• A minimum of around 60% or 6.0 CGPA through academics.
• No active backlog at the time of the process, with gap rules that vary by drive.

Package tiers, per candidate reports across recent drives, sit at three levels:

Track	Reported package	Typical trigger
Analyst	around 4.25 LPA	clears the process, coding not met
Analyst Star	around 5.75 LPA	solves one coding problem
Senior Analyst	around 7.50 LPA	solves both coding problems

You do not choose your track on the application. The process routes you based on scores. That is why the coding-heavy stages carry weight beyond a simple pass.

Common mistakes candidates make

Most avoidable losses come from preparing for the wrong test. A few patterns repeat every drive:

• Studying only the old pseudocode-and-quant format and skipping the AI Literacy section.
• Treating the AI-assisted coding stage like a normal coding round, then freezing when it asks for step-by-step direction.
• Practising code writing but never debugging, so the debugging stage feels unfamiliar.
• Assuming a strong technical score offsets a weak stage. It does not, because the eliminatory stages stand alone.
• Reading each game’s instructions after the timer starts, and running out of time in the cognitive stage.

How to prepare for the Exceller framework

Sequence your prep to match the stages. Start with the technical module. It is both eliminatory and the widest in scope, so revise DSA basics, core CS fundamentals, and add an AI Literacy layer covering generative-AI concepts and prompt structure. Build debugging reps by reading and fixing broken C, C++, and Java snippets, not only by writing new programs. Practise the AI-assisted flow by framing a problem in clear steps before asking for any code. Keep spoken English warm with short daily speaking drills. Treat the cognitive games as a timing exercise rather than something to memorise.

A realistic order for a three-to-four week window: technical module first, then debugging, then the AI-assisted flow, with communication and cognitive practice woven through as daily short reps. Front-load the eliminatory stages. The package-tier stages reward you later, but only if you are still in the process.

If you want the older-format context for how Capgemini structured its aptitude rounds before this framework, the discussion of Capgemini aptitude questions is a useful companion, and the game-based guide above is the natural next read once you have mapped your weak stage. Start with whichever eliminatory stage you are least sure of, because that is the one most likely to end your process early.