CDS Cutoff: Academy-wise Written and Final History, 2012–2025

A CDS cutoff is meaningful only when you attach four details: the examination sitting, the academy or course, the selection stage and the marks available at that stage. A written cutoff of 98 for OTA and a written cutoff of 124 for IMA cannot be compared as raw numbers because the written maxima differ. The final cutoff belongs to the combined selection score; it is not a second written-paper threshold. Keeping these distinctions clear makes the tables useful for preparation instead of turning them into misleading targets.

This guide transcribes the official UPSC PDFs linked from the current cutoff directory and its archive. We found 27 sittings, beginning with CDS-II 2012 and ending with CDS-II 2025. Both sittings are included for every year from 2013 to 2025. The reviewed archive did not provide CDS-I 2012 or earlier CDS sittings, so this is a documented available-history series, not a claim to cover every CDS exam ever held. Each sitting has a direct official source below.

UPSC’s 2025-II PDF gives a minimum of 20% in each subject. It reports decimal aggregate cutoffs, which this guide preserves rather than rounding to whole marks. The OTA men’s written cutoff is 80.01; the OTA women’s is 98.06. Their final cutoffs also differ. Do not copy the men’s row into the women’s course or use the shorthand “OTA cutoff” without identifying the course. The official document is the authority for those figures.

The last candidate at the written stage and the last recommended candidate need not be the same person. Consequently, final cutoff minus written cutoff does not reveal an official SSB minimum. A candidate can enter SSB with a written score higher than the written threshold; the final aggregate then reflects that candidate’s own written and interview marks. Medical fitness, eligibility, preferences and the other entry conditions still apply. A cutoff table is a record of marks at a stage, not a certificate of admission.

Read the newest rows first, then look down the same academy column. “I” and “II” identify separate examinations, not two stages of one exam. Each number is an official written aggregate cutoff in marks. IMA, INA and AFA columns are out of 300; the two OTA columns are out of 200. The chart uses percentages of each academy’s maximum for bar width, but the table retains the original marks so you can audit the conversion.

The final table uses the same sitting order and preserves separate OTA men’s and women’s courses. IMA, INA and AFA are out of 600, while OTA is out of 400. These marks describe the last recommended candidate, using the terminology in UPSC’s documents. A larger final number does not automatically indicate a more difficult written paper because the final stage also includes SSB and has a different denominator.

Many rows in this history use 20% in each paper, but that is not a universal historical rule. The 2012-II document records 12 marks in each subject. The 2013-I document states 15% in each subject, and also attaches a 40% SSB condition to its final figures. CDS-I 2019 records 20% in Paper I, 10% in Paper II and 20% in Paper III. CDS-I 2024 states 18% in each subject. These footnotes are part of the official record, not details to replace with a standard coaching summary.

For a future sitting, consult the applicable notification and the eventual official cutoff publication. A lower historical subject minimum does not entitle an applicant to the same minimum later. When checking a mock score against an older row, write the footnote beside the aggregate. This prevents a false sense of qualification from a high English score hiding a weak GK score. The subject-minimum table describes historical requirements; it does not announce future qualifying marks.

Our arithmetic: IMA’s written cutoff moves from 127 in 2023-II to 117 in 2024-II and 124.11 in 2025-II. OTA men moves from 104 to 81 to 80.01 across those same sittings; OTA women moves from 104 to 81 to 98.06. These sequences demonstrate why a single rising-or-falling headline is inadequate. The two OTA courses can have different outcomes, and comparing only one sitting per year excludes the other examination’s evidence.

The cutoff PDFs do not by themselves prove why a cutoff changed. Paper difficulty, applicant performance and vacancies are possible factors to investigate with separate evidence; this table does not estimate their individual effect. It also does not measure an individual’s selection probability. Historical minimum and maximum values describe the range in the collected series. They are not a confidence interval for the next exam, and they should not be labelled as a statistically guaranteed safe score.

Our preparation method: take a timed official past paper under its printed instructions, calculate net marks after the stated penalty and record the result paper by paper. Compare only with the same written maximum and the relevant academy/course. Then separate two questions: does each paper clear the historical subject minimum, and how does the total compare with the historical aggregate? Passing one of those comparisons does not imply passing the other.

The examples are invented score scenarios for explanation, not invented exam questions or candidate results. They show how the same total can produce different comparisons depending on the course and subject marks. Keep several timed attempts in your log rather than using the best attempt as your baseline. Include the paper name, date, testing conditions and error categories. A repeated weakness is more actionable than a one-off total above an old cutoff.

Choose a personal preparation buffer from the variability in your own timed scores and the historical series, but label it as your plan. There is no official “safe score” in these PDFs. If your mock total swings widely, improve consistency before relying on a buffer. Review skipped questions, wrong answers and time lost separately. A strategy that increases attempts while increasing negative marks may look energetic without producing a better net result.

The reviewed 2026-II notification states that a wrong answer loses one third of the marks assigned to that question, a multiple response counts as wrong and a blank response receives no penalty. Use the question’s assigned marks when calculating; do not assume every paper awards one mark per item. If an official answer key is pending, keep your score estimate provisional. An unofficial answer key can help discussion, but it cannot turn an estimate into UPSC’s recorded mark.

The cutoff explorer intentionally accepts no personal answers or speculative score input. Its job is to make the documented series easier to inspect. For scoring your own paper, use the original paper instructions and the applicable official key. Do not round decimal cutoffs down when deciding whether a displayed score is above a past threshold. Retain the precision provided by the official source, especially in the 2025-II rows.

UPSC’s 2024-I and 2024-II cutoff documents describe age seniority first for equal aggregate marks, then written marks, then GK marks if needed. The 2025-I document lists written total first, then GK marks, then age. This difference is another reason to open the exact PDF rather than applying the newest-looking rule to every year. The guide’s history tables do not attempt to reproduce every sitting’s complete tie procedure; those procedures remain in the linked documents.

Open UPSC’s cutoff directory, locate the exact examination title and read the academy row together with its footnotes. If the sitting has moved, use the archive. Check whether the PDF is revised; the 2014-I link in this series is the revised document supplied by the archive. Save the examination label as well as the download date. A filename can contain an upload date or even a misleading year fragment, so the examination heading inside the PDF is the identifier used for this dataset.

At the 7 October 2026 check, neither reviewed directory listed a 2026 CDS cutoff. An expected cutoff published by a coaching site is a prediction unless UPSC has issued that figure. We therefore leave 2026 out of the historical tables. When an official new sitting is added, it should receive its own source row and subject-minimum check; existing historical values should change only if UPSC publishes a correction.