CMI does not publish official cutoffs or percentiles. Discover the data-backed safe score for CMI MSc Data Science 2027, derived from 293 previous year questions, the 100-mark exam pattern, and precise topic weightage analysis.
Last verified: October 7, 2026
CMI does not publish official cutoff marks or percentiles for the MSc Data Science entrance exam. However, based on the 100-mark exam pattern and historical difficulty, a raw score of 65 to 75 is widely considered a safe target to secure an interview call. This target relies heavily on mastering the School Level Mathematics section, which accounts for 37.2 percent of all previous year questions.
CMI does not publish official cutoff marks or percentiles. However, based on the 100-mark exam pattern and historical difficulty, a raw score of 65-75 is widely considered a safe target to secure an interview call, heavily reliant on mastering the 37.2% School Level Mathematics weightage.
The core academic requirement for this program is an undergraduate degree. This includes B.A., B.Sc., B.Math., B.Stat., B.E., B.Tech. or an equivalent, provided you have a background in Mathematics, Statistics or Computer Science. This wording is deliberately broader than a strict Statistics or CS degree requirement, as it actively admits engineers, physicists and other quantitative graduates. Final-year undergraduates who expect to complete their degree by the start of the relevant academic year are also eligible to apply.
Because the institute does not release percentile data, candidates must benchmark their preparation against raw score targets derived from the exam structure. The following matrix maps historical performance patterns to interview probability.
| Metric | Raw Score | Net Correct | Note |
|---|---|---|---|
| High Probability Interview Call | 65-75 / 100 | ~15/20 Part A + ~10-12/20 Part B (with partial credit) | CMI does not publish official percentiles or cutoffs. This range is derived from the 100-mark structure (40 Part A + 60 Part B) and historical applicant pool difficulty. |
| Moderate Probability Interview Call | 50-64 / 100 | ~12/20 Part A + ~8-10/20 Part B (with partial credit) | Varies significantly by year-wise paper difficulty. Strong performance in the 37.2% weightage School Math section is critical to remain in this bracket. |
| Low Probability / Waitlist | 35-49 / 100 | ~8/20 Part A + ~6-8/20 Part B | Typically insufficient for the General category, but may clear the relaxed qualifying scores applied for reserved categories (SC/ST/OBC-NCL/EWS/PC). |
No. CMI explicitly states that it does not compute or publish an all-India rank, percentile, or category-wise cutoff list for the MSc Data Science entrance exam. Instead, selected candidates are listed solely by their application ID on the official results page.
Cold Reality: Most candidates waste hours chasing a mythical cutoff mark that the institute does not publish. The only metric you control is your raw score in Part A and your methodical working in Part B. Stop guessing percentiles and start securing the 37.2 percent weightage in School Level Mathematics.
The exam consists of 40 questions totaling 100 marks. Part A contains 20 objective questions worth 2 marks each (40 marks total) with no partial credit. Part B contains 20 descriptive questions worth 3 marks each (60 marks total), rewarding partial credit.
The MSc Data Science admission test is a separate, dedicated paper. It is not shared with any other entrance exams and has run as its own distinct question paper every year since the program first intake in 2018. Official past papers and solutions are published for 2018 through 2026.
Two features make it different from other tests at the institute. First is the timing. While all exams are held on the same afternoon, the MSc Data Science paper runs from 2:00 PM to 5:30 PM. This is half an hour longer than the 2:00 PM to 5:00 PM window given to MSc and PhD Mathematics, Computer Science, and Physics papers. This extra time reflects the larger descriptive component of the Data Science paper. The exam is held offline in pen-and-paper mode at roughly 37 exam centres across India.
What most candidates get wrong here is treating all topics equally. Based on our analysis of 293 previous year questions, you must prioritize high-yield areas and avoid time sinks during the actual exam.
Highest yield. Forms the foundation of Part A objective questions where accuracy is mandatory due to zero partial credit.
Second highest weightage. Heavily tested in both Part A and Part B descriptive sections with compound problem structures.
Critical for logical reasoning and combinatorics. Frequently appears as multi-step Part B questions that reward partial credit.
High time cost for tracing loops or array processing. Easy to make off-by-one errors under strict time pressure.
Can devolve into lengthy manual calculations if a modular arithmetic shortcut or theorem is not immediately visible.
Extremely low yield. Only 2 questions in the entire 293-PYQ dataset, making it a poor return on investment for initial study time.
Minimal appearance as a distinct combined unit. Better to focus on core Probability Theory first.
Planning your preparation requires awareness of the official timeline. While exact dates for the 2027 cycle are pending official notification, historical patterns provide a reliable projection.
Prepare your undergraduate degree transcripts and category certificates. The official portal will open for new registrations soon.
The exam is scheduled for 2:00 PM to 5:30 PM. Ensure you have your admit card and valid government ID proof ready.
| Milestone | Projected Date (IST) | Action Required |
|---|---|---|
| Application Window Opens | March 1, 2027 | Register on the official portal and select the dedicated MSc Data Science paper. |
| Entrance Examination | May 15, 2027 (2:00 PM) | Appear for the 3.5 hour offline pen-and-paper test at your assigned center. |
| Results and Interview Shortlist | Early June 2027 | Check the official results page. Selected candidates are listed by application ID only. |
For official updates, always refer to the Chennai Mathematical Institute Admissions Page.
Generic preparation leaves you guessing which topics actually matter. MastersUp builds each learner a personalized, AI-driven study plan. Machine learning tracks your real performance topic by topic, spots weak and strong areas, and adjusts what you practice next. Every practice question is curated for your specific gaps, not generic random practice.
MastersUp's focus-mode flow allows for learning on the go, one topic at a time, eliminating distraction and building deep conceptual clarity.
The system tracks your prep window: 6 full revisions across 12 months, 5 across 8 months, 4 across 6 months, 3 across 4 months, 2 across 3 months, or a compact sprint mode when time is short.
The flow is simple. Log in, pick your exam, and MastersUp separates your weak and strong topics. You continue with the default plan or customize it, then you start learning. Every weak spot gets surfaced as you go. Even at 1 hour of daily practice, you can see exactly where you stand, topic by topic, against other students on the platform.
Since CMI does not publish official cutoffs, a safe score is an estimate. Based on the 100-mark pattern and historical difficulty, scoring 65+ raw marks (e.g., 15/20 in Part A and 10/20 in Part B with partial credit) is generally considered a strong target to secure an interview call.
No. CMI's official results page explicitly states that it does not compute or publish an all-India rank, percentile, or category-wise cutoff list. Admitted candidates are listed solely by their application ID, making it fundamentally different from exams like JEE or GATE.
While there is no fixed qualifying mark, candidates typically need to clear a relaxed qualifying score for reserved categories (SC/ST/OBC-NCL/EWS/PC) and a higher implicit threshold for the General category. Maximizing Part A accuracy and showing method in Part B is the only verified strategy.