Compete

Coding

International Olympiad in AI

International Olympiad in Artificial Intelligence (IOAI)

The newest international science olympiad, with a scientific round and a practical machine learning round.

Run by IOAI, with national organising bodies

Official website How to prepare Reviews and tips

Key facts

Eligible grades
High school students living in the country they represent
IOAI's own rule is 'high school students'. The US national qualifier, USAAIO, is broader and accepts K-12 students in the US and Canada, though only those meeting IOAI's rule can go on to the international event.
Source, checked 2026-09-01
Recommended preparation
Programming plus an understanding of machine learning fundamentals
Why: The international syllabus covers machine learning, neural networks and model training. The US qualifier accepts K-12 students, so preparation rather than year of study determines who is ready.
Based on the official syllabus and content covered.
This is Compete's reading of the competition's documented content, not a requirement stated by the organiser. You may enter without it.
Compete guidance, not an eligibility restriction. Students outside this are still eligible if the official eligibility above allows them.
Eligible ages
Less than 20 years old on the first day of the international competition
Stated by IOAI and repeated by USAAIO, which adds that contestants must not be full-time university students in the competition year.
Source, checked 2026-09-01
Individual or team
Team · Each country is represented by a national team of four contestants and a team leader, selected by the country coordinator from national competitions in AI or related subjects. Contestants compete individually at IOAI; the Team Challenge does not affect medals.
Where it runs
Online and in person · International
Entry cost
National qualifiers are usually free; international travel is often fundedNot confirmed. The USAAIO contest rules page did not state a fee on the version checked.
Registration deadline
For US and Canadian students: USAAIO registration for the 2027 edition is open and closes 31 January 2027 at 11:59 p.m. EST. 173 days leftThere is no direct entry to IOAI. Students in other countries must find their own National Olympiad in AI; deadlines vary by country.
Competition dates
IOAI 2027, the 4th edition, will be held in Singapore from 4 to 10 July 2027, jointly hosted by AI Singapore and Nanyang Technological University. IOAI 2026 was held in Astana, Kazakhstan, 2 to 8 August 2026.
Difficulty
EliteSelection is measured in the top fraction of a percent nationally. Typically reached by qualifying through a lower round rather than entered directly.
Prizes
Medals and international recognitionMedals at IOAI. The 2027 medal boundaries and any national awards were not confirmed on the pages checked.

What International Olympiad in AI is

The International Olympiad in Artificial Intelligence is the newest of the international science olympiads, first held in 2024. National teams compete in two rounds: a scientific round of open-ended AI problems solved in teams over a long window, and a practical round of hands-on machine learning tasks. Countries select teams through their own national qualifiers, and the United States runs a national AI olympiad for this purpose.

How the competition works

Each country runs a national selection process, typically an online qualifying exam followed by a training and selection camp. National teams of four students then attend the international olympiad, where they work on a scientific round of open-ended problems and a practical round involving real machine learning tasks with provided compute.

Eligibility notes

You must qualify through your own country's national selection. Entry rules and qualifier formats are set nationally.

Is International Olympiad in AI right for you?

Experience needed
Advanced. You need real Python, comfort with linear algebra and probability, and hands-on experience training models.
Time commitment
A year or more of serious self-directed study.
What you actually do
The scientific round gives open-ended AI problems worked in teams over an extended window. The practical round gives hands-on tasks with provided compute, closer to a supervised machine learning contest.

A good fit if

  • You already train models rather than only reading about them
  • You have strong maths and want to apply it
  • You are aiming at the highest international level

Probably not for you if

  • Students new to programming, who should start with USACO or a project competition
  • Students who want a solo competition

Compare this against competitions matched to you

Common mistakes

  • Studying AI theory without ever training a model. The practical round is hands-on.
  • Neglecting the mathematics. Linear algebra and probability underpin both rounds.
  • Assuming the international event is the entry point. Your national qualifier is the actual gate and its deadline is what matters.

How to register

  1. Find your country's national AI olympiad organiser through the international site.
  2. Enter the national qualifying round, which is usually an online exam covering machine learning fundamentals and mathematics.
  3. Qualifiers attend national training and a selection process.
  4. Selected students form the national team of four.

Register on the official site

What skills you need

  • Python and PyTorch or similar frameworks
  • Linear algebra and probability
  • Neural network architectures
  • Prompt and model evaluation
  • Reading machine learning papers
  • Team problem solving

How to prepare

IOAI preparation is genuine machine learning practice rather than exam cramming. The strongest preparation is building and training real models, reading the standard textbooks, and working competitive machine learning problems, because the practical round is closer to a Kaggle-style task than to a written exam.

What to study

Mathematics
Linear algebra, probability, calculus for backpropagation, and optimisation. Both rounds assume all of it.
Core machine learning
Regression, classification, regularisation, cross-validation, bias and variance. The unglamorous foundation that decides practical round scores.
Deep learning
Convolutional and transformer architectures, training dynamics, fine-tuning, evaluation. Implement these yourself rather than only using libraries.
Practical tooling
Python, PyTorch, data loading, GPU use, and debugging a training run that is not converging.
Reading papers
The scientific round poses open-ended problems that reward familiarity with current literature.

Full preparation guide, including a week-by-week plan

Recommended resources

Past problems and materials

Only materials the organiser or an authorised archive publishes openly are listed here.

Helpful videos

Frequently asked questions

How do I enter?

Through your country's national AI olympiad qualifier. There is no direct entry to the international event. Find your national organiser through the IOAI site, since formats and deadlines differ considerably between countries.

Do I need my own GPU?

Not to prepare. Free tiers on Google Colab and Kaggle provide enough compute for learning and for most practice projects, and the international rounds provide compute.

Reviews, tips and resources from other students

Posts are public. Do not share your full name, school, address or contact details.

Your rating
0 / 1200
Sort

Something out of date?

If a deadline, fee or eligibility rule on this page no longer matches the official site, flag it and it goes into the review queue.

You might also look at

StructureDivisions by grade or level
  • FormatIndividual
  • WhereOnline
    • Check site
    Coding WAICY A project-based global AI competition with tracks for art and video as well as technical builds.
    Check site
    Coding Codeforces Free short-format algorithmic contests several times a month, open to anyone with a rating attached.
    Check site
    Science Physics Olympiad (F=ma) A 25-question mechanics exam that opens the pathway to the United States physics team.
    Check site