The Indian National AI Olympiad (INAIO) 2027 holds its Stage 1 test on Sunday, 13 December 2026, from 10 am to 12 pm. It is an in-person test and AI assistants are not allowed. The olympiad is open to students in Classes 9 to 12 from any board. It is India’s official route to the International Olympiad in Artificial Intelligence (IOAI), which will be held in Singapore from 4 to 10 July 2027. School registration for Stage 1 is open now on inaio.in, so if you are interested you have about seven weeks to prepare.
Key Facts: INAIO 2027 at a Glance
- Organiser: ACM India, the Indian arm of the Association for Computing Machinery (inaio.in).
- Who can apply: Indian citizens studying in India in Classes 9 to 12, from any board, born after 1 July 2007.
- Stage 1: 13 December 2026, 10 am to 12 pm, in person. No AI assistants are allowed.
- Stage 2: 24 January 2027, online. AI assistants are allowed in some sections.
- National camp: May 2027, in person at IIIT Hyderabad.
- International final: IOAI 2027, 4 to 10 July 2027, Singapore, hosted by AI Singapore and NTU Singapore (NTU, 10 Sep 2026).
- India’s record: At IOAI 2026, India won 1 gold, 5 silver, 1 bronze and 1 honourable mention. At its IOAI debut in 2025, India placed 3rd on the medal table (inaio.in).
- Registration: Through your school for Stage 1. ACM India has not published the 2027 fee on its homepage. The 2025 cycle charged Rs 767 including GST (ACM India).
- Contact: inaio@acmindia.org
What Is the Indian National AI Olympiad?
The Indian National AI Olympiad is a multi-stage contest that ACM India uses to pick the students who represent India at the IOAI. It works like the better-known Maths and Informatics Olympiads. A broad first test is followed by tougher rounds, then coaching, then a camp where the final team is chosen.
The international contest it feeds into is still new. According to its public record, IOAI began in Burgas, Bulgaria in 2024 with nearly 200 students. It moved to Beijing in 2025 with about 310 students from 63 countries. In 2026 it was held in Astana, Kazakhstan, with 466 contestants from 108 countries and territories.
One detail is worth knowing. When ACM India announced INAIO 2026 in December 2025, IOAI 2026 was planned for Abu Dhabi (The Hans India, 20 Dec 2025). In March 2026 the organisers moved it to Astana for safety reasons tied to the situation in the Middle East (BES Media, 21 Mar 2026). Older articles still mention Abu Dhabi, which is why you should take dates and venues from the official site and not from aggregator pages.
Who Is Eligible for INAIO 2027?
You are eligible for INAIO 2027 if you are an Indian citizen studying in India in Class 9, 10, 11 or 12 and were born after 1 July 2007. Any board is fine: CBSE, ICSE, a state board or IB. The official page lists two exclusions:
- Students who have already finished Class 12 cannot take part. This includes students on a gap year.
- Students planning to retake Class 12 board exams are also not eligible.
The age cut-off comes from the IOAI rules, which are open to secondary-school students under 20. Each country may send up to two teams of up to four students.
Direct Entry to Stage 2
Some students skip Stage 1 because they have already done well in a related olympiad. For the 2027 cycle, inaio.in lists these groups:
- INAIO 2026 Training Camp participants.
- INOI 2026 and INOI 2027 medallists (Indian National Olympiad in Informatics).
- PLO 2026 Training Camp participants (Panini Linguistics Olympiad).
- Students who qualified for INMO 2025-26 or 2026-27 through RMO (Maths Olympiad).
- Bebras India top scorers, in the Mahavira and Ramanujam categories only.
If you already compete in computing olympiads, the Zonal Informatics Olympiad (ZIO-2027) is on 21 November 2026, and registration closes on 6 November (IARCS). Doing well in the informatics track is a second way into the AI olympiad pipeline. Most students, though, will enter through Stage 1.
INAIO 2027 Stages and Dates
INAIO 2027 has five stages, which run from December 2026 to June 2027, followed by the international final in July. The schedule below is from inaio.in.
- Stage 1, 13 December 2026: A two-hour in-person test with no AI assistants. This is the broad screening round.
- Stage 2, 24 January 2027: An online test where AI assistants are allowed in some sections. Expect Python and machine-learning problems here.
- Stage 3, March to April 2027: Online coaching and evaluation for selected students.
- Stage 4, May 2027: An in-person national training camp at IIIT Hyderabad.
- Stage 5, June 2027: Team preparation for the students chosen for IOAI.
- IOAI 2027, 4 to 10 July 2027: The international final in Singapore.
For Class 10 and Class 12 students, Stage 2 and the camp overlap with pre-board exams and board season. Plan for that now. Do not wait until January to find out you have a clash.
What Does the Stage 1 Test Look Like?
The official Stage 1 format for 2027 has not been posted on the INAIO homepage yet. A student prep repository says the format note is expected about two weeks before the exam (GitHub, checked 16 Oct 2026). That is a third-party source, so treat it as a guide only.
The best clue is the previous published format. In the 2025 cycle, ACM India’s page described Stage 1 like this:
- 80 marks, plus up to 20 bonus marks.
- Ten short questions worth 2 marks each, about 2 minutes each.
- Four multi-part questions worth 15 marks each, about 15 minutes each.
- Answers given as multiple-choice options or integers from 0 to 99.
The 2025 Stage 1 was online and lasted 90 minutes. The 2027 Stage 1 is in person, two hours long and bans AI assistants, so the format may have changed. What is unlikely to change is the type of thinking it tests: careful reasoning about data, probability and how models behave, all done without a computer doing the work for you.
INAIO Syllabus: What to Study
The Stage 1 and Stage 2 syllabus is a linked document on inaio.in. Stages 3 and 4 follow the IOAI 2027 syllabus. Read the official document first. The areas below are the ones ACM India listed for the 2025 cycle, and they match the IOAI syllabus.
Foundations (Stage 1 focus)
- Python: loops, functions, lists and dictionaries, and reading code to predict its output.
- Probability and statistics: conditional probability, expectation, mean, variance and distributions.
- Linear algebra: vectors, matrices, dot products and matrix multiplication.
- Optimisation: what a loss function is, and how gradient descent reduces it step by step.
Core Machine Learning (Stages 1 and 2)
- Supervised learning: linear and logistic regression, k-nearest neighbours, decision trees and SVMs.
- Unsupervised learning: k-means clustering and PCA.
- Evaluation: train/test splits, overfitting and underfitting, confusion matrices, precision and recall, and ROC curves.
Advanced Topics (Later Stages)
The IOAI 2026 syllabus goes much further. It covers neural networks and backpropagation, CNNs, transformers and attention, pre-trained models like BERT and CLIP, diffusion models, and audio models like Whisper. On the practical side it expects NumPy, Pandas, scikit-learn and PyTorch. At its May 2026 community meeting, IOAI said the syllabus is being widened to include more multimodal AI (IOAI, May 2026). You do not need all of this for Stage 1. You will need it if you reach the camp.
An 8-Week INAIO Preparation Plan
From 21 October to 13 December you have just under eight weeks. The plan below assumes about 5 to 7 hours a week alongside school. Adjust it to your level. A student who already codes in Python can move faster through the first two weeks.
- Weeks 1-2 (to 3 Nov): Python and maths basics. Write small programs by hand and predict their output before you run them. Revise probability from your school maths, then learn vectors and matrix multiplication.
- Weeks 3-4 (to 17 Nov): Classical ML ideas. Learn what linear regression, k-NN, decision trees and k-means actually compute. Work one small example of each on paper, with five or six data points.
- Week 5 (to 24 Nov): Model evaluation. Practise computing accuracy, precision and recall from a confusion matrix. Practise spotting overfitting from training and test numbers.
- Week 6 (to 1 Dec): Neural network intuition. Understand a single neuron, activation functions and one step of gradient descent. Do a few calculations by hand.
- Week 7 (to 8 Dec): Past papers under exam conditions. Use official past papers if inaio.in or ACM India provides them. Time yourself strictly, and work on paper without AI tools, the way Stage 1 will be.
- Week 8 (exam week): Review your errors only. Re-solve the questions you got wrong. Read the official format note once it is out. Sleep properly the two nights before the test.
Free Places to Learn
You do not need a paid coaching course for Stage 1. These free resources cover most of the foundation syllabus:
- Kaggle Learn: short, free courses on Python, Pandas and introductory machine learning, with exercises in the browser.
- Khan Academy: probability, statistics and linear algebra, explained from the basics.
- 3Blue1Brown’s neural networks series on YouTube: a visual explanation of how networks learn and what backpropagation does.
- scikit-learn user guide: the official documentation explains each classical algorithm, with examples. It is useful for Stage 2.
How to Use AI While Preparing
Use AI as a tutor, not as an answer machine, because Stage 1 bans AI assistants. If a chatbot solves every practice problem for you, you will arrive on 13 December without the skill the test measures. Ask it to explain a concept, give you a fresh practice question or check your reasoning after you have tried the problem yourself.
Stage 2 is different. AI assistants are allowed for some sections there, and in the 2025 cycle ACM India allowed free ChatGPT and Gemini in Stage 2 (ACM India). If you reach Stage 2, practise solving ML coding tasks with an assistant and checking its output. Spotting when an AI-generated model or answer is wrong is a skill in its own right.
Is the AI Olympiad Worth It If You Don’t Win?
Yes. For most students the value is the learning, not a medal. Only a handful of students reach the national camp, and fewer travel to IOAI. Two months of focused work on probability, Python and machine learning still gives you a real head start for a computer science, data science or engineering degree.
It also helps beyond the score. For INAIO 2026, ACM India said all participants would get digital certificates and the top three national rankers would get prizes (The Hans India). Confirm whether the same applies for 2027. A project or notebook you build while preparing can also go on a college application or portfolio.
Keep it in proportion if you are in Class 10 or 12. Board exams matter more for your next step. If the olympiad starts to feel like extra pressure, talk to a teacher, parent or school counsellor about how much time to give it.
Frequently Asked Questions
Q. When is the Indian National AI Olympiad 2027 exam?
Ans. Stage 1 of the Indian National AI Olympiad 2027 is on Sunday, 13 December 2026, from 10 am to 12 pm. It is an in-person test. Stage 2 follows online on 24 January 2027.
Q. Can a Class 9 student take the AI Olympiad?
Ans. Yes. INAIO 2027 is open to Indian students in Classes 9 to 12 from any board who were born after 1 July 2007. Students who have finished Class 12 are not eligible.
Q. How do I register for INAIO 2027?
Ans. Stage 1 registration is done through schools, and school registration is open on inaio.in. Ask your school’s computer science teacher or olympiad coordinator to register. If you have problems, email inaio@acmindia.org.
Q. Do I need to know coding for the AI Olympiad?
Ans. Basic Python helps from Stage 1, because questions can test code reasoning. Stage 2 and later stages involve hands-on machine-learning coding. Strong maths, especially probability and linear algebra, matters as much as coding.
Q. How did India do at the International Olympiad in AI?
Ans. India won 1 gold, 5 silver and 1 bronze medal, plus an honourable mention, at IOAI 2026 in Astana. At its debut in Beijing in 2025, India placed third on the medal table with 3 golds, 2 silvers and 1 bronze.
Next Steps
Ask your school this week whether it has registered for the Indian National AI Olympiad 2027 Stage 1. If it hasn’t, point the coordinator to inaio.in. Then start the first two weeks of the plan above, with Python and probability. ACM India is expected to publish the fee, the test centres and the official Stage 1 format closer to December, so check inaio.in regularly. We will update this guide when those details are out.