1.4.1Introduction to the Indian National AI Olympiad

What is INAIO?

The Indian National AI Olympiad (INAIO) is the national selection process for the students who will represent India at IOAI. It is organized by ACM India in partnership with ACM IKDD.

INAIO does more than select the final team. Its later stages also train the shortlisted students. You begin with broad tests of mathematics, logic, and AI. As you move ahead, the tests become more practical and much closer to the work you will do at IOAI.

Who could enter INAIO 2026?

Stage 1 was open to Indian citizens studying in India in Grades 9 to 12, from any school board, provided they were born after August 5, 2006. Students who had already completed Class 12 were not eligible. Students from any grade could directly enter Stage 2 if they met one of the listed conditions based on achievements in other Olympiads.

The five stages

The Olympiad mainly has five stages:

1. Stage 1 - Mathematical and logical reasoning: Mathematics-grounded questions test your logic, theoretical understanding, and basic ability to manage time. For INAIO 2026, this was a two-hour online test with objective questions only, and AI assistants were not allowed.
2. Stage 2 - Theory and practical implementation: This stage mixes objective questions, long-form problems, and hands-on modelling. Its level is approximately that of the Bronze and Silver modules in this guide. AI assistants may be allowed for some sections, depending on that year's rules.
3. Stage 3 - Online coaching and evaluation: Around 30–40 shortlisted students attend online training sessions and lectures. Their progress is tracked through further objective, long-form, and hands-on evaluations. The required knowledge is similar to Stage 2, with roughly a solid Silver-level base expected.
4. Stage 4 - The national training camp: Around 20–30 students attend an offline camp at IIIT Hyderabad. It has lectures, lots of practice tests, tests that mimic proper exam conditions, and technical interviews. These assessments determine the final eight students.
5. Stage 5 - Team preparation: The eight selected students receive further training and check-ups before IOAI.

After Stage 5, the students participate in IOAI as the two Indian teams. The exact number of students shortlisted at each earlier stage can change from year to year.

The approximate timeline

The full process takes place over several months. For INAIO 2026, the official schedule looked roughly like this:

This timeline is only a rough guide. The dates may shift earlier or later in another year. If IOAI itself is held earlier, the INAIO stages will also need to move earlier.

Stage 1: mathematics and reasoning

Stage 1 is mainly mathematical and theoretical, so you need to know the theory and reason with it quickly. You also need good time management because the paper moves across several different ideas.

If you are eligible through achievements in another Olympiad, you may be allowed to skip Stage 1 and directly take Stage 2. For INAIO 2026, the following students received direct entry:

• Participants in the INAIO 2025 Training Camp.
• INOI 2025 and INOI 2026 medallists.
• Participants in the PLO 2025 Training Camp.
• Students who qualified to take INMO 2024–25 or INMO 2025–26 based on their RMO performance.
• The highest scorers in the Mahavira and Ramanujam categories of the Bebras India Computing Challenge.

INMO 2026 qualifiers were also allowed to skip Stage 1 because INMO and INAIO Stage 1 were held on the same date.

What should you know for Stage 1?

The exact paper changes every year. Still, the official syllabus and previous paper give a good idea of the kind of thinking that appears. You should be comfortable with:

• pattern recognition, propositional logic, constraints, and careful case analysis;
• basic programming constructs such as variables, operations, conditions, and loops;
• functions, polynomials, quadratic equations, series, summations, and common inequalities;
• probability, conditional probability, Bayes' theorem, and descriptive statistics;
• matrices, basic matrix operations, and systems of linear equations;
• simple maxima, minima, and linear optimization under constraints;
• reading tables, hierarchies, graphs, plots, and other data representations;
• the train-test process, loss functions, and common regression and classification metrics; and
• the basic reasoning behind algorithms such as linear regression, K-NN, decision trees, K-means, PCA, Naive Bayes, PageRank, and TF-IDF.

You do not need deep knowledge of every algorithm before opening the paper. The syllabus says that questions on these algorithms will provide the relevant background. Previous papers have also introduced ideas such as convolution and word embeddings inside the question itself. You must be able to understand that explanation, connect it to the mathematics you know, and reason from it.

For Stage 1, mathematics and logical reasoning should be enough to get you through, given that you know a reasonable amount of the mathematics this guide needs as a prerequisite.

Stage 2: theory meets implementation

Stage 2 is relatively more similar to the format followed over the next stages. It mixes objective questions, long-form problems, and hands-on modelling. Theory and practical implementation have similar weight, and each section may have its own cutoff that you need to pass. Knowing only the mathematics is no longer enough. You must also work with data and implement methods in code.

For Stage 2, the main addition is the practical part. The Bronze and Silver sections of this guide should give you enough coverage for this stage. Other than the theory required for Stage 1, you need approximately a Silver-level understanding of classical machine learning and deep learning. You should also be able to write basic Python, use NumPy, Pandas, Matplotlib, Seaborn, and scikit-learn, implement the listed methods, and explore a dataset properly.

Stages 3 to 5

For Stage 3, the needed topics are similar to Stage 2. The focus shifts towards learning through the online lectures, completing the evaluations, and showing that you can keep up with the expected level.

The next jump occurs at the training camp. The competition becomes stronger, the tests come closer to real exam conditions, and you need at least Gold-level skills. For Stage 4, you should be thorough with the Gold modules across all the main domains rather than depending on only one strong area.

Stage 5 begins after the final eight students have been selected. At this point, the aim is team preparation: fixing remaining gaps, practising under contest conditions, and getting ready for the exact IOAI format.